PASS1B QC and Analysis report: Proteomics
David Amar (Stanford), Pierre Jean Beltran (Broad), James Sanford (PNNL), David Jimenez-Morales (Stanford)
2021/09/14
In this document we analyze the PASS1B proteomics animal data of MoTrPAC. The flow steps are: (1) load the data from the google cloud bucket, (2) preprocess each dataset, (3) PCA analysis and correlation with metadata variables, and (4) flagging potential outliers, (5) removing outliers as revised by the proteomics groups, (6) perform differential abundance analysis.
# CONFIGURATIONS
# Specify the parameters required for the analysis
# This directory should have the MotrpacRatTraining6moQCRep repository
# clone from: https://github.com/MoTrPAC/MotrpacRatTraining6moQCRep
repo_local_dir = "~/github/MoTrPAC/MotrpacRatTraining6moQCRep"
# Runnable command for gsutil
gsutil_cmd = "~/bin/google-cloud-sdk/bin/gsutil"
# These paths were removed for security reasons, see the repo's readme
bucket = "ADDPATH"
# Where should the data be downloaded to
download_from_bucket = FALSE # If FALSE, it assumes the data is there
local_data_dir = "ADDPATH"
# where to print the data and dea tables
upload_to_bucket = FALSE # TRUE upload the data to the bucket
out_bucket = "ADDPATH"
dea_folder_name = "dea"
suffix = "20210914.txt"
# Specify bucket and local path for the phenotypic data
pheno_bucket = "ADDPATH"
pheno_local_dir = "ADDPATH"
# specify parameters for filtering by missing values
max_na_freq = 0.67
# Should PCA use imputed data?
pca_using_impute=FALSE
# Specify how many PCs to examine
num_pcs = 5
num_pcs_for_outlier_analysis = 3
# Specify the significance level for association analysis
# (e.g., between a PC and a clinical variable)
p_thr = 0.0001Specify the pipeline, metadata, and sample variables for the analysis.
# Define technical and biological variables to be considered in the QC
pipeline_qc_cols = c("tmt_plex","num_NA")
biospec_cols = c("registration.sex",
"key.anirandgroup",
"registration.batchnumber",
"training.staffid",
"is_control",
"vo2.max.test.vo2_max_visit1", # this assumes that visit1's are aligned
"terminal.weight.mg",
"time_to_freeze",
"timepoint",
"bid",
"pid")# load functions and libraries
source(paste0(repo_local_dir,"/tools/unsupervised_normalization_functions.R"))
source(paste0(repo_local_dir,"/tools/gcp_functions.R"))
source(paste0(repo_local_dir,"/tools/qc_report_aux_functions.R"))
source(paste0(repo_local_dir,"/tools/config_session.R"))
source(paste0(repo_local_dir,"/tools/association_analysis_methods.R"))
source(paste0(repo_local_dir,"/tools/ptm_protein_correction.R"))1. Load data
1.1 Download data
if(download_from_bucket){
obj = download_bucket_to_local_dir(bucket,
local_path=path.expand(local_data_dir),
remove_prev_files = TRUE,
GSUTIL_PATH=gsutil_cmd)
}1.2 Parse proteomics results
The code belows take the metadata and ration tables. For raw intensities (rii) - specify the matrix regex in “parse_proteomics_datasets_from_download_obj”.
obj = list(
downloaded_files = list.files(local_data_dir,full.names = TRUE,recursive = TRUE),
local_path = local_data_dir
)
proteomics_data = parse_proteomics_datasets_from_download_obj(
obj,local_path=local_data_dir,remove_prev_files = TRUE,GSUTIL_PATH=gsutil_cmd,
platforms = c("prot-ph","prot-pr","prot-ac", "prot-ub")
)## [1] "--------------------------------------------------------------------"
## [1] " analyzing dataset: t53-cortex,prot-ph"
## [1] "Using the annotation columns in the data file: motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/proteomics/t53-cortex/prot-ph/motrpac_pass1b-06_t53-cortex_prot-ph_ratio-results.txt"
## [1] "Column names in the row annotation data:"
## [1] "protein_id" "gene_symbol" "entrez_id"
## [4] "redundant_ids" "organism_name" "is_contaminant"
## [7] "ptm_id" "confident_score" "confident_site"
## [10] "flanking_sequence" "ptm_score"
## [1] "using ptm_id as the feature name"
## [1] "--------------------------------------------------------------------"
## [1] " analyzing dataset: t53-cortex,prot-pr"
## [1] "Using the annotation columns in the data file: motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/proteomics/t53-cortex/prot-pr/motrpac_pass1b-06_t53-cortex_prot-pr_ratio-results.txt"
## [1] "Column names in the row annotation data:"
## [1] "protein_id" "gene_symbol" "entrez_id" "redundant_ids"
## [5] "organism_name" "is_contaminant" "num_peptides" "percent_coverage"
## [9] "protein_score"
## [1] "using protein id as feature name"
## [1] "--------------------------------------------------------------------"
## [1] " analyzing dataset: t55-gastrocnemius,prot-ph"
## [1] "Using the annotation columns in the data file: motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/proteomics/t55-gastrocnemius/prot-ph/motrpac_pass1b-06_t55-gastrocnemius_prot-ph_ratio-results.txt"
## [1] "Column names in the row annotation data:"
## [1] "protein_id" "gene_symbol" "entrez_id"
## [4] "redundant_ids" "organism_name" "is_contaminant"
## [7] "ptm_id" "confident_score" "confident_site"
## [10] "flanking_sequence" "ptm_score"
## [1] "using ptm_id as the feature name"
## [1] "--------------------------------------------------------------------"
## [1] " analyzing dataset: t55-gastrocnemius,prot-pr"
## [1] "Using the annotation columns in the data file: motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/proteomics/t55-gastrocnemius/prot-pr/motrpac_pass1b-06_t55-gastrocnemius_prot-pr_ratio-results.txt"
## [1] "Column names in the row annotation data:"
## [1] "protein_id" "gene_symbol" "entrez_id" "redundant_ids"
## [5] "organism_name" "is_contaminant" "num_peptides" "percent_coverage"
## [9] "protein_score"
## [1] "using protein id as feature name"
## [1] "--------------------------------------------------------------------"
## [1] " analyzing dataset: t58-heart,prot-ph"
## [1] "Using the annotation columns in the data file: motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/proteomics/t58-heart/prot-ph/motrpac_pass1b-06_t58-heart_prot-ph_ratio-results.txt"
## [1] "Column names in the row annotation data:"
## [1] "protein_id" "gene_symbol" "entrez_id"
## [4] "redundant_ids" "organism_name" "is_contaminant"
## [7] "ptm_id" "confident_score" "confident_site"
## [10] "flanking_sequence" "ptm_score"
## [1] "using ptm_id as the feature name"
## [1] "--------------------------------------------------------------------"
## [1] " analyzing dataset: t58-heart,prot-pr"
## [1] "Using the annotation columns in the data file: motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/proteomics/t58-heart/prot-pr/motrpac_pass1b-06_t58-heart_prot-pr_ratio-results.txt"
## [1] "Column names in the row annotation data:"
## [1] "protein_id" "gene_symbol" "entrez_id" "redundant_ids"
## [5] "organism_name" "is_contaminant" "num_peptides" "percent_coverage"
## [9] "protein_score"
## [1] "using protein id as feature name"
## [1] "--------------------------------------------------------------------"
## [1] " analyzing dataset: t58-heart,prot-ac"
## [1] "Using the annotation columns in the data file: motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/proteomics/t58-heart/prot-ac/motrpac_pass1b-06_t58-heart_prot-ac_ratio-results.txt"
## [1] "Column names in the row annotation data:"
## [1] "protein_id" "gene_symbol" "entrez_id"
## [4] "redundant_ids" "organism_name" "is_contaminant"
## [7] "ptm_id" "confident_score" "confident_site"
## [10] "flanking_sequence" "ptm_score"
## [1] "using ptm_id as the feature name"
## [1] "--------------------------------------------------------------------"
## [1] " analyzing dataset: t58-heart,prot-ub"
## [1] "Using the annotation columns in the data file: motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/proteomics/t58-heart/prot-ub/motrpac_pass1b-06_t58-heart_prot-ub_ratio-results.txt"
## [1] "Column names in the row annotation data:"
## [1] "protein_id" "gene_symbol" "entrez_id"
## [4] "redundant_ids" "organism_name" "is_contaminant"
## [7] "ptm_id" "confident_score" "confident_site"
## [10] "flanking_sequence" "ptm_score"
## [1] "using ptm_id as the feature name"
## [1] "--------------------------------------------------------------------"
## [1] " analyzing dataset: t59-kidney,prot-ph"
## [1] "Using the annotation columns in the data file: motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/proteomics/t59-kidney/prot-ph/motrpac_pass1b-06_t59-kidney_prot-ph_ratio-results.txt"
## [1] "Column names in the row annotation data:"
## [1] "protein_id" "gene_symbol" "entrez_id"
## [4] "redundant_ids" "organism_name" "is_contaminant"
## [7] "ptm_id" "confident_score" "confident_site"
## [10] "flanking_sequence" "ptm_score"
## [1] "using ptm_id as the feature name"
## [1] "--------------------------------------------------------------------"
## [1] " analyzing dataset: t59-kidney,prot-pr"
## [1] "Using the annotation columns in the data file: motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/proteomics/t59-kidney/prot-pr/motrpac_pass1b-06_t59-kidney_prot-pr_ratio-results.txt"
## [1] "Column names in the row annotation data:"
## [1] "protein_id" "gene_symbol" "entrez_id" "redundant_ids"
## [5] "organism_name" "is_contaminant" "num_peptides" "percent_coverage"
## [9] "protein_score"
## [1] "using protein id as feature name"
## [1] "--------------------------------------------------------------------"
## [1] " analyzing dataset: t66-lung,prot-ph"
## [1] "Using the annotation columns in the data file: motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/proteomics/t66-lung/prot-ph/motrpac_pass1b-06_t66-lung_prot-ph_ratio-results.txt"
## [1] "Column names in the row annotation data:"
## [1] "protein_id" "gene_symbol" "entrez_id"
## [4] "redundant_ids" "organism_name" "is_contaminant"
## [7] "ptm_id" "confident_score" "confident_site"
## [10] "flanking_sequence" "ptm_score"
## [1] "using ptm_id as the feature name"
## [1] "--------------------------------------------------------------------"
## [1] " analyzing dataset: t66-lung,prot-pr"
## [1] "Using the annotation columns in the data file: motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/proteomics/t66-lung/prot-pr/motrpac_pass1b-06_t66-lung_prot-pr_ratio-results.txt"
## [1] "Column names in the row annotation data:"
## [1] "protein_id" "gene_symbol" "entrez_id" "redundant_ids"
## [5] "organism_name" "is_contaminant" "num_peptides" "percent_coverage"
## [9] "protein_score"
## [1] "using protein id as feature name"
## [1] "--------------------------------------------------------------------"
## [1] " analyzing dataset: t68-liver,prot-ph"
## [1] "Using the annotation columns in the data file: motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/proteomics/t68-liver/prot-ph/motrpac_pass1b-06_t68-liver_prot-ph_ratio-results.txt"
## [1] "Column names in the row annotation data:"
## [1] "protein_id" "gene_symbol" "entrez_id"
## [4] "redundant_ids" "organism_name" "is_contaminant"
## [7] "ptm_id" "confident_score" "confident_site"
## [10] "flanking_sequence" "ptm_score"
## [1] "using ptm_id as the feature name"
## [1] "--------------------------------------------------------------------"
## [1] " analyzing dataset: t68-liver,prot-pr"
## [1] "Using the annotation columns in the data file: motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/proteomics/t68-liver/prot-pr/motrpac_pass1b-06_t68-liver_prot-pr_ratio-results.txt"
## [1] "Column names in the row annotation data:"
## [1] "protein_id" "gene_symbol" "entrez_id" "redundant_ids"
## [5] "organism_name" "is_contaminant" "num_peptides" "percent_coverage"
## [9] "protein_score"
## [1] "using protein id as feature name"
## [1] "--------------------------------------------------------------------"
## [1] " analyzing dataset: t68-liver,prot-ac"
## [1] "Using the annotation columns in the data file: motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/proteomics/t68-liver/prot-ac/motrpac_pass1b-06_t68-liver_prot-ac_ratio-results.txt"
## [1] "Column names in the row annotation data:"
## [1] "protein_id" "gene_symbol" "entrez_id"
## [4] "redundant_ids" "organism_name" "is_contaminant"
## [7] "ptm_id" "confident_score" "confident_site"
## [10] "flanking_sequence" "ptm_score"
## [1] "using ptm_id as the feature name"
## [1] "--------------------------------------------------------------------"
## [1] " analyzing dataset: t68-liver,prot-ub"
## [1] "Using the annotation columns in the data file: motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/proteomics/t68-liver/prot-ub/motrpac_pass1b-06_t68-liver_prot-ub_ratio-results.txt"
## [1] "Column names in the row annotation data:"
## [1] "protein_id" "gene_symbol" "entrez_id"
## [4] "redundant_ids" "organism_name" "is_contaminant"
## [7] "ptm_id" "confident_score" "confident_site"
## [10] "flanking_sequence" "ptm_score"
## [1] "using ptm_id as the feature name"
## [1] "--------------------------------------------------------------------"
## [1] " analyzing dataset: t70-white-adipose,prot-ph"
## [1] "Using the annotation columns in the data file: motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/proteomics/t70-white-adipose/prot-ph/motrpac_pass1b-06_t70-white-adipose_prot-ph_ratio-results.txt"
## [1] "Column names in the row annotation data:"
## [1] "protein_id" "gene_symbol" "entrez_id"
## [4] "redundant_ids" "organism_name" "is_contaminant"
## [7] "ptm_id" "confident_score" "confident_site"
## [10] "flanking_sequence" "ptm_score"
## [1] "using ptm_id as the feature name"
## [1] "--------------------------------------------------------------------"
## [1] " analyzing dataset: t70-white-adipose,prot-pr"
## [1] "Using the annotation columns in the data file: motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/proteomics/t70-white-adipose/prot-pr/motrpac_pass1b-06_t70-white-adipose_prot-pr_ratio-results.txt"
## [1] "Column names in the row annotation data:"
## [1] "protein_id" "gene_symbol" "entrez_id" "redundant_ids"
## [5] "organism_name" "is_contaminant" "num_peptides" "percent_coverage"
## [9] "protein_score"
## [1] "using protein id as feature name"failed_datasets = proteomics_data$failed_datasets
proteomics_data = proteomics_data$proteomics_parsed_datasets
#this is commented out: the printed buckets here will refer to the "ac" datasets,
# and these were not submitted for this release.
if(length(failed_datasets)>1){
print("Assays/data not available (or failed):")
local_fname = file.path(local_data_dir, paste0("log-pass1b-06_datasets-not-processed_",suffix))
write.table(failed_datasets, local_fname, row.names = FALSE,quote=FALSE,sep="\t",col.names = FALSE)
}## [1] "Assays/data not available (or failed):"1.3 Download and parse phenotypic data
# Add the pheno data
pheno_data = parse_pheno_data(pheno_bucket,
local_path = path.expand(pheno_local_dir),
remove_prev_files = TRUE,
GSUTIL_PATH=gsutil_cmd)
# add a tissue variable (for convinience)
pheno_data$viallabel_data$tissue =
pheno_data$viallabel_data$specimen.processing.sampletypedescription
# add the time to freeze variable ((for convinience))
pheno_data$viallabel_data$time_to_freeze =
pheno_data$viallabel_data$calculated.variables.frozetime_after_train -
pheno_data$viallabel_data$calculated.variables.deathtime_after_train
# some freeze times are negative because the tissues were taken
# before sacrifice time, zero these cases
pheno_data$viallabel_data$time_to_freeze = pmax(0,pheno_data$viallabel_data$time_to_freeze)
# add a binary is_control variable
pheno_data$viallabel_data$is_control = as.numeric(grepl("control",
pheno_data$viallabel_data$key.anirandgroup,ignore.case = T))
# add the timepoint as a number
# x - the text description of the group
get_numeric_timepoint<-function(x){
v = rep(0,length(x))
tps = c("Eight"=8,"Four"=4,"One"=1,"Two"=2)
for(tp in names(tps)){
v[grepl(paste0(tp,"-week"),x,perl = T,ignore.case = T)] = tps[tp]
}
return(v)
}
pheno_data$viallabel_data$timepoint = get_numeric_timepoint(
pheno_data$viallabel_data$key.anirandgroup
)2 Normalization and processing
2.1 Normalize and filter the data
- Median mad normalization
- Remove batch effect
- Impute (for PCA)
Note: We use limma’s removeBatchEffect to account for plex variation. We also keep a version of the data with imputed missing values. The new data objects are saved into the proteomics_data list.
for(dataset_name in names(proteomics_data)){
print(paste("processing",dataset_name))
curr_data = proteomics_data[[dataset_name]]$sample_data
# median value normalization
sample_meta = proteomics_data[[dataset_name]]$sample_meta[colnames(curr_data),]
# Exclude features with many NAs
row_NA_data = rowSums(is.na(curr_data))/ncol(curr_data)
feature_inds = row_NA_data < max_na_freq
# filter and normalize
norm_data = as.matrix(curr_data[feature_inds,])
norm_data = run_median_mad_norm(norm_data)
# b for batch corrected
norm_data_b = limma::removeBatchEffect(norm_data,sample_meta$tmt_plex)
# Impute - use capture.output to avoid prints
# This will be used for PCA
capture.output({
norm_data_b_imp = impute::impute.knn(as.matrix(norm_data_b))$data
norm_data_imp = impute::impute.knn(as.matrix(norm_data))$data
},file=NULL)
# Median normalization (per MOP)
# norm_data_imp = run_median_mad_norm(norm_data_imp)
if(any(is.na(norm_data_imp))){
print(paste("Imputation failed in:",dataset_name))
}
proteomics_data[[dataset_name]][["norm_filtered"]] = list(
norm_data_b = norm_data_b, # normalized and batch corrected
norm_data_b_imp = norm_data_b_imp, # normalizes, batch corrected and imputed
norm_data = norm_data, # normalized only, no batch correction, no imputation
norm_data_imp = norm_data_imp, # normalized with imputation, no batch correction
feature_inds = feature_inds
)
print(paste0('done, objects saved into proteomics_data[["',
dataset_name,'"]][[,norm_filtered]]'))
}2.2 Examine normalization effects
Examine the effect of the normalization process, which includes removing features with many NAs. For each dataset show the boxplots of three randomly selected plexes.
for(i in 1:length(proteomics_data)){
unnorm_data = proteomics_data[[i]]$sample_data
norm_data = proteomics_data[[i]][["norm_filtered"]][[1]]
plex = proteomics_data[[i]]$sample_meta[colnames(norm_data),"tmt_plex"]
plex = as.factor(plex)
# order the data by the plexes
ord = order(plex)
unnorm_data = unnorm_data[,ord]
norm_data = norm_data[,ord]
plex = plex[ord]
# take a subset of the plexes
samp_plex = sample(unique(plex))[1:3]
samp = which(plex %in% samp_plex)
# plot
curr_name = names(proteomics_data)[i]
colnames(unnorm_data) = NULL;colnames(norm_data) = NULL
boxplot(unnorm_data[,samp],
main=paste0(curr_name,"\nraw data (",nrow(unnorm_data)," features)"),
ylab = "log2 ratio",xaxt="n",col=plex)
boxplot(norm_data[,samp],
main=paste0(curr_name,"\nnorm data (ratios) (",nrow(norm_data)," features)"),
ylab = "log2 ratio",xaxt="n",col=plex)
}2.3 PTM Protein-abundance correction
Correct PTM datasets with protein abundances. Most relevant for ubiquitylome data.
for(ptm_assay in c("prot-ub","prot-ac","prot-ph")){
#Select the PTM assay that has to be corrected
ptm_data_names <- grep(ptm_assay,names(proteomics_data),value = T)
tissues <- str_extract(ptm_data_names,"(\\w|-)+")
#Loop through all tissues in this PTM assay and perform protein correction
#Protein corrected dataset is saved under norm_filtered$norm_data_b_protein_corrected
for(tissue in tissues){
ptm_name <- paste(tissue,ptm_assay,sep=",")
prot_name <- paste(tissue,"prot-pr",sep=",")
cat("Protein correction - Processing dataset",paste(tissue,ptm_assay,sep=","),"...")
#Identify matching protein id and save to row_annot in the PTM dataset
proteomics_data[[ptm_name]]$row_annot <-
ptm_protein_match(pr_data = proteomics_data[[prot_name]]$norm_filtered$norm_data_b,
ptm_row_annot = proteomics_data[[ptm_name]]$row_annot,
prot_row_annot = proteomics_data[[prot_name]]$row_annot)
#Perform protein correction of PTM data
proteomics_data[[ptm_name]]$norm_filtered$norm_data_b_protein_corrected <-
ptm_protein_correction(ptm_data = proteomics_data[[ptm_name]]$norm_filtered$norm_data_b,
pr_data = proteomics_data[[prot_name]]$norm_filtered$norm_data_b,
ptm_row_annot = proteomics_data[[ptm_name]]$row_annot,
prot_row_annot = proteomics_data[[prot_name]]$row_annot,
plot_graphs = T)
#Remove sites with many missing values as indicated by feature_inds
proteomics_data[[ptm_name]]$norm_filtered$norm_data_b_protein_corrected <-
proteomics_data[[ptm_name]]$norm_filtered$norm_data_b_protein_corrected[
proteomics_data[[ptm_name]]$norm_filtered$feature_inds,]
}
}## Protein correction - Processing dataset t58-heart,prot-ub ...[1] "508991 points with proteome match out of 536760 (94.8%)."
## [1] "Fitting model..."
## [1] "Success."
##
## Call:
## lm(formula = value.ptm ~ value.prot, data = data, na.action = na.exclude)
##
## Residuals:
## Min 1Q Median 3Q Max
## -6.1175 -0.2021 -0.0101 0.1891 3.7684
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.0061779 0.0006305 9.799 <2e-16 ***
## value.prot 0.2503471 0.0052074 48.075 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3578 on 322107 degrees of freedom
## (214651 observations deleted due to missingness)
## Multiple R-squared: 0.007124, Adjusted R-squared: 0.007121
## F-statistic: 2311 on 1 and 322107 DF, p-value: < 2.2e-16
## Protein correction - Processing dataset t68-liver,prot-ub ...[1] "717459 points with proteome match out of 749400 (95.7%)."
## [1] "Fitting model..."
## [1] "Success."
##
## Call:
## lm(formula = value.ptm ~ value.prot, data = data, na.action = na.exclude)
##
## Residuals:
## Min 1Q Median 3Q Max
## -4.6426 -0.2045 -0.0010 0.2039 3.8543
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.0037322 0.0006096 6.123 9.22e-10 ***
## value.prot 0.4603204 0.0017321 265.764 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3794 on 388692 degrees of freedom
## (360706 observations deleted due to missingness)
## Multiple R-squared: 0.1538, Adjusted R-squared: 0.1538
## F-statistic: 7.063e+04 on 1 and 388692 DF, p-value: < 2.2e-16
## Protein correction - Processing dataset t58-heart,prot-ac ...[1] "330018 points with proteome match out of 345060 (95.6%)."
## [1] "Fitting model..."
## [1] "Success."
##
## Call:
## lm(formula = value.ptm ~ value.prot, data = data, na.action = na.exclude)
##
## Residuals:
## Min 1Q Median 3Q Max
## -6.3194 -0.1441 -0.0130 0.1299 6.2150
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 0.0037316 0.0007151 5.218 1.81e-07 ***
## value.prot 0.3775431 0.0050331 75.012 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3281 on 210856 degrees of freedom
## (134202 observations deleted due to missingness)
## Multiple R-squared: 0.02599, Adjusted R-squared: 0.02599
## F-statistic: 5627 on 1 and 210856 DF, p-value: < 2.2e-16
## Protein correction - Processing dataset t68-liver,prot-ac ...[1] "625382 points with proteome match out of 645354 (96.9%)."
## [1] "Fitting model..."
## [1] "Success."
##
## Call:
## lm(formula = value.ptm ~ value.prot, data = data, na.action = na.exclude)
##
## Residuals:
## Min 1Q Median 3Q Max
## -6.6010 -0.1683 -0.0005 0.1648 4.9365
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) -0.0008948 0.0005488 -1.631 0.103
## value.prot 0.6439628 0.0020414 315.459 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3479 on 403020 degrees of freedom
## (242332 observations deleted due to missingness)
## Multiple R-squared: 0.198, Adjusted R-squared: 0.198
## F-statistic: 9.951e+04 on 1 and 403020 DF, p-value: < 2.2e-16
## Protein correction - Processing dataset t53-cortex,prot-ph ...[1] "3616301 points with proteome match out of 4069260 (88.9%)."
## [1] "Fitting model..."
## [1] "Success."
##
## Call:
## lm(formula = value.ptm ~ value.prot, data = data, na.action = na.exclude)
##
## Residuals:
## Min 1Q Median 3Q Max
## -9.5722 -0.1199 0.0055 0.1278 6.4964
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) -0.0062784 0.0001909 -32.89 <2e-16 ***
## value.prot 0.1072188 0.0014144 75.81 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2515 on 1735626 degrees of freedom
## (2333632 observations deleted due to missingness)
## Multiple R-squared: 0.0033, Adjusted R-squared: 0.003299
## F-statistic: 5746 on 1 and 1735626 DF, p-value: < 2.2e-16
## Protein correction - Processing dataset t55-gastrocnemius,prot-ph ...[1] "2097186 points with proteome match out of 2606340 (80.5%)."
## [1] "Fitting model..."
## [1] "Success."
##
## Call:
## lm(formula = value.ptm ~ value.prot, data = data, na.action = na.exclude)
##
## Residuals:
## Min 1Q Median 3Q Max
## -7.4267 -0.2236 0.0067 0.2339 4.6641
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) -0.0148050 0.0004559 -32.48 <2e-16 ***
## value.prot 0.2157740 0.0017208 125.39 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.4657 on 1048828 degrees of freedom
## (1557510 observations deleted due to missingness)
## Multiple R-squared: 0.01477, Adjusted R-squared: 0.01477
## F-statistic: 1.572e+04 on 1 and 1048828 DF, p-value: < 2.2e-16
## Protein correction - Processing dataset t58-heart,prot-ph ...[1] "2466530 points with proteome match out of 2906880 (84.9%)."
## [1] "Fitting model..."
## [1] "Success."
##
## Call:
## lm(formula = value.ptm ~ value.prot, data = data, na.action = na.exclude)
##
## Residuals:
## Min 1Q Median 3Q Max
## -9.4843 -0.1565 0.0008 0.1580 5.3862
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) -0.002969 0.000254 -11.69 <2e-16 ***
## value.prot 0.208000 0.001678 123.99 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3192 on 1580107 degrees of freedom
## (1326771 observations deleted due to missingness)
## Multiple R-squared: 0.009635, Adjusted R-squared: 0.009635
## F-statistic: 1.537e+04 on 1 and 1580107 DF, p-value: < 2.2e-16
## Protein correction - Processing dataset t59-kidney,prot-ph ...[1] "2415147 points with proteome match out of 2771340 (87.1%)."
## [1] "Fitting model..."
## [1] "Success."
##
## Call:
## lm(formula = value.ptm ~ value.prot, data = data, na.action = na.exclude)
##
## Residuals:
## Min 1Q Median 3Q Max
## -6.5221 -0.1379 0.0024 0.1404 4.0082
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) -0.0039386 0.0002571 -15.32 <2e-16 ***
## value.prot 0.2094322 0.0018962 110.45 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.2658 on 1069265 degrees of freedom
## (1702073 observations deleted due to missingness)
## Multiple R-squared: 0.01128, Adjusted R-squared: 0.01128
## F-statistic: 1.22e+04 on 1 and 1069265 DF, p-value: < 2.2e-16
## Protein correction - Processing dataset t66-lung,prot-ph ...[1] "3609313 points with proteome match out of 4024980 (89.7%)."
## [1] "Fitting model..."
## [1] "Success."
##
## Call:
## lm(formula = value.ptm ~ value.prot, data = data, na.action = na.exclude)
##
## Residuals:
## Min 1Q Median 3Q Max
## -9.5629 -0.1655 0.0125 0.1805 3.8742
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) -0.0130596 0.0002311 -56.5 <2e-16 ***
## value.prot 0.1839376 0.0013530 135.9 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.315 on 1857487 degrees of freedom
## (2167491 observations deleted due to missingness)
## Multiple R-squared: 0.009852, Adjusted R-squared: 0.009851
## F-statistic: 1.848e+04 on 1 and 1857487 DF, p-value: < 2.2e-16
## Protein correction - Processing dataset t68-liver,prot-ph ...[1] "2726429 points with proteome match out of 3168240 (86.1%)."
## [1] "Fitting model..."
## [1] "Success."
##
## Call:
## lm(formula = value.ptm ~ value.prot, data = data, na.action = na.exclude)
##
## Residuals:
## Min 1Q Median 3Q Max
## -9.2918 -0.1742 0.0063 0.1814 5.6259
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) -0.0051438 0.0002705 -19.02 <2e-16 ***
## value.prot 0.3876859 0.0012781 303.33 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.3536 on 1708681 degrees of freedom
## (1459557 observations deleted due to missingness)
## Multiple R-squared: 0.0511, Adjusted R-squared: 0.0511
## F-statistic: 9.201e+04 on 1 and 1708681 DF, p-value: < 2.2e-16
## Protein correction - Processing dataset t70-white-adipose,prot-ph ...[1] "2263871 points with proteome match out of 2644860 (85.6%)."
## [1] "Fitting model..."
## [1] "Success."
##
## Call:
## lm(formula = value.ptm ~ value.prot, data = data, na.action = na.exclude)
##
## Residuals:
## Min 1Q Median 3Q Max
## -10.6080 -0.2169 0.0114 0.2356 5.9893
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) -0.0034578 0.0004456 -7.759 8.55e-15 ***
## value.prot 0.6013227 0.0008119 740.671 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.4597 on 1073525 degrees of freedom
## (1571333 observations deleted due to missingness)
## Multiple R-squared: 0.3382, Adjusted R-squared: 0.3382
## F-statistic: 5.486e+05 on 1 and 1073525 DF, p-value: < 2.2e-16
3. Quality control
3.1 PCA before and after adjusting for batch
Run PCA analysis on the imputed data, with and without mitigating batch effects:
for(dataset_name in names(proteomics_data)){
data_obj = proteomics_data[[dataset_name]]
data_samples = colnames(data_obj[["norm_filtered"]][["norm_data_b"]])
randgroup = pheno_data$viallabel_data[data_samples,"key.anirandgroup"]
sex = pheno_data$viallabel_data[data_samples,"registration.sex"]
sex[sex=="1"] = "F";sex[sex=="2"]="M"
plex = data_obj$sample_meta[data_samples,"tmt_plex"]
# pca after removing batch effects
if(pca_using_impute){
curr_data = data_obj[["norm_filtered"]][["norm_data_b_imp"]]
}
else{
curr_data = data_obj[["norm_filtered"]][["norm_data_b"]]
curr_data = curr_data[rowSums(is.na(curr_data))==0,]
}
curr_pca = prcomp(t(curr_data),scale. = T,center = T)
curr_pcax = curr_pca$x[,1:num_pcs]
explained_var = summary(curr_pca)[["importance"]][3,5]
df = data.frame(curr_pcax,sex=sex,plex=plex,randgroup = randgroup)
proteomics_data[[dataset_name]][["norm_filtered"]][["pca_b"]] = list(
pcax = df,explained_var=explained_var
)
# pca without removing batch effects
if(pca_using_impute){
curr_data = data_obj[["norm_filtered"]][["norm_data_imp"]]
}
else{
curr_data = data_obj[["norm_filtered"]][["norm_data"]]
curr_data = curr_data[rowSums(is.na(curr_data))==0,]
}
curr_pca = prcomp(t(curr_data),scale. = T,center = T)
curr_pcax = curr_pca$x[,1:num_pcs]
explained_var = summary(curr_pca)[["importance"]][3,5]
df = data.frame(curr_pcax,sex=sex,plex=plex,randgroup = randgroup)
proteomics_data[[dataset_name]][["norm_filtered"]][["pca_w_b"]] = list(
pcax = df,explained_var=explained_var
)
}PCA plots:
for(dataset_name in names(proteomics_data)){
df = proteomics_data[[dataset_name]][["norm_filtered"]][["pca_w_b"]]$pcax
p = ggplot(df) +
geom_point(aes(x=PC1, y=PC2,col=plex, shape=sex)) +
ggtitle(paste(dataset_name,"before batch effects analysis",sep=": "))
plot(p)
df = proteomics_data[[dataset_name]][["norm_filtered"]][["pca_b"]]$pcax
p = ggplot(df) +
geom_point(aes(x=PC1, y=PC2,col=plex, shape=sex)) +
ggtitle(paste(dataset_name,"after batch effects analysis",sep=": "))
plot(p)
}
3.2 PCA visualization of sex and group
We plot the top two PCs for each tissue, color and shape correspond to randomization group and sex.
for(dataset_name in names(proteomics_data)){
df = proteomics_data[[dataset_name]][["norm_filtered"]][["pca_b"]]$pcax
p = ggplot(df) +
geom_point(aes(x=PC1, y=PC2,col=randgroup, shape=sex)) +
ggtitle(dataset_name)
plot(p)
}
3.3 Correlations with clinical variables
Here we analyze the top principal components (5) in each tissue and compute their association with the selected variables above (e.g., the pipeline qc metrics). We use Spearman correlation and a linear test for significance. For each dataset we also add the correlations among the metadata variables. These two analyses should be interpreted together, as the analyzed variables are not independent.
pcs_vs_qc_var_report = c()
metadata_variable_assoc_report = c()
for(dataset_name in names(proteomics_data)){
# Get the projected PCs
curr_pcax = proteomics_data[[dataset_name]][["norm_filtered"]][["pca_b"]][[1]]
curr_pcax = curr_pcax[,grepl("^PC",names(curr_pcax))]
curr_pipeline_meta = proteomics_data[[dataset_name]]$sample_meta
curr_data_samples = rownames(curr_pcax)
curr_meta = pheno_data$viallabel_data[curr_data_samples,biospec_cols]
# remove metadata variables with NAs
curr_meta = curr_meta[,!apply(is.na(curr_meta),2,any)]
# remove bid and pid
curr_meta = curr_meta[,!grepl("pid|bid",names(curr_meta))]
# Add batch
curr_meta$plex = as.numeric(
as.factor(curr_pipeline_meta[curr_data_samples,"tmt_plex"]))
curr_meta_numeric_cols = sapply(curr_meta,is.numeric)
# Remove zero sd columns
curr_meta_numeric_cols[curr_meta_numeric_cols] =
apply(curr_meta[,curr_meta_numeric_cols],2,sd)>0
corrs = cor(curr_pcax,curr_meta[,curr_meta_numeric_cols],method="spearman")
corrsp = pairwise_eval(
curr_pcax,curr_meta[,curr_meta_numeric_cols],func=pairwise_association_wrapper,
f=1,max_num_vals_for_discrete=8)
# Some ggplots have to be printed to be shown in the notebook
print(ggcorrplot(t(corrs),lab=TRUE,lab_size=2.5,hc.order = F) +
ggtitle(dataset_name) +
theme(plot.title = element_text(hjust = 0.5,size=20)))
for(i in 1:nrow(corrsp)){
for(j in 1:ncol(corrsp)){
if(corrsp[i,j]>p_thr){next}
pcs_vs_qc_var_report = rbind(pcs_vs_qc_var_report,
c(dataset_name,
rownames(corrsp)[i],colnames(corrsp)[j],corrs[i,j],corrsp[i,j])
)
}
}
colnames(pcs_vs_qc_var_report) = c("Dataset(tissue,site)","PC",
"qc_metric","rho(spearman)","p-value")
####
# compute correlations among the metadata variables
####
corrs = cor(curr_meta[,curr_meta_numeric_cols],method="spearman")
corrsp = pairwise_eval(
curr_meta[,curr_meta_numeric_cols],func=pairwise_association_wrapper,
f=1,max_num_vals_for_discrete=8)
# Some ggplots have to be printed to be shown in the notebook
print(ggcorrplot(corrs,lab=TRUE,lab_size=2.5,hc.order = T))
for(n1 in rownames(corrsp)){
for(n2 in rownames(corrsp)){
if(n1==n2){break}
if(n1 %in% biospec_cols &&
n2 %in% biospec_cols) {next}
if(corrsp[n1,n2]>p_thr){next}
metadata_variable_assoc_report = rbind(metadata_variable_assoc_report,
c(dataset_name,n1,n2,corrs[n1,n2],corrsp[n1,n2])
)
}
}
if(!is.null(dim(metadata_variable_assoc_report))){
colnames(metadata_variable_assoc_report) = c(
"Dataset(tissue,site)","Var1","Var2","rho(spearman)","p-value")
}
}
# Format the reports - for a nicer presentation in a table
pcs_vs_qc_var_report[,5] = format(
as.numeric(pcs_vs_qc_var_report[,5]),digits=3)
pcs_vs_qc_var_report[,4] = format(
as.numeric(pcs_vs_qc_var_report[,4]),digits=3)
metadata_variable_assoc_report[,5] = format(
as.numeric(metadata_variable_assoc_report[,5]),digits=3)
metadata_variable_assoc_report[,4] = format(
as.numeric(metadata_variable_assoc_report[,4]),digits=3)3.3.1 Summary of significant clinical variables
Table of clinical variables with significant correlations to the data
kable(pcs_vs_qc_var_report,longtable=TRUE,
caption = "Correlations between PCs and other variables") %>%
kable_styling(latex_options = c("repeat_header"),font_size = 8)| Dataset(tissue,site) | PC | qc_metric | rho(spearman) | p-value |
|---|---|---|---|---|
| t53-cortex,prot-ph | PC1 | time_to_freeze | -0.582 | 2.98e-07 |
| t55-gastrocnemius,prot-ph | PC3 | registration.sex | 0.703 | 2.58e-09 |
| t55-gastrocnemius,prot-ph | PC3 | training.staffid | -0.551 | 1.86e-05 |
| t55-gastrocnemius,prot-ph | PC3 | terminal.weight.mg | 0.587 | 1.17e-07 |
| t55-gastrocnemius,prot-pr | PC1 | registration.sex | -0.656 | 2.76e-08 |
| t55-gastrocnemius,prot-pr | PC1 | training.staffid | 0.618 | 1.64e-07 |
| t55-gastrocnemius,prot-pr | PC1 | terminal.weight.mg | -0.501 | 4.60e-07 |
| t55-gastrocnemius,prot-pr | PC5 | vo2.max.test.vo2_max_visit1 | -0.426 | 3.67e-05 |
| t58-heart,prot-ph | PC1 | registration.sex | -0.853 | 6.44e-15 |
| t58-heart,prot-ph | PC1 | training.staffid | 0.606 | 2.15e-08 |
| t58-heart,prot-ph | PC1 | terminal.weight.mg | -0.748 | 1.80e-13 |
| t58-heart,prot-pr | PC1 | registration.sex | -0.866 | 1.29e-28 |
| t58-heart,prot-pr | PC1 | training.staffid | 0.520 | 6.68e-06 |
| t58-heart,prot-pr | PC1 | vo2.max.test.vo2_max_visit1 | 0.559 | 2.92e-06 |
| t58-heart,prot-pr | PC1 | terminal.weight.mg | -0.739 | 7.38e-22 |
| t58-heart,prot-ac | PC1 | registration.sex | 0.598 | 2.00e-06 |
| t58-heart,prot-ac | PC1 | terminal.weight.mg | 0.579 | 9.29e-06 |
| t58-heart,prot-ac | PC2 | registration.sex | -0.562 | 9.30e-06 |
| t58-heart,prot-ac | PC2 | terminal.weight.mg | -0.483 | 4.53e-06 |
| t58-heart,prot-ac | PC3 | is_control | 0.524 | 5.15e-05 |
| t58-heart,prot-ac | PC5 | is_control | 0.544 | 3.48e-05 |
| t58-heart,prot-ub | PC1 | registration.sex | 0.676 | 1.18e-06 |
| t58-heart,prot-ub | PC1 | terminal.weight.mg | 0.558 | 7.50e-06 |
| t58-heart,prot-ub | PC3 | is_control | 0.505 | 3.22e-06 |
| t59-kidney,prot-ph | PC1 | registration.sex | -0.860 | 5.03e-26 |
| t59-kidney,prot-ph | PC1 | vo2.max.test.vo2_max_visit1 | 0.594 | 2.89e-07 |
| t59-kidney,prot-ph | PC1 | terminal.weight.mg | -0.772 | 2.66e-22 |
| t59-kidney,prot-pr | PC1 | registration.sex | -0.866 | 1.21e-55 |
| t59-kidney,prot-pr | PC1 | training.staffid | 0.472 | 4.07e-05 |
| t59-kidney,prot-pr | PC1 | vo2.max.test.vo2_max_visit1 | 0.553 | 4.73e-07 |
| t59-kidney,prot-pr | PC1 | terminal.weight.mg | -0.779 | 4.09e-31 |
| t66-lung,prot-ph | PC2 | registration.sex | 0.776 | 4.44e-12 |
| t66-lung,prot-ph | PC2 | training.staffid | -0.553 | 4.15e-07 |
| t66-lung,prot-ph | PC2 | vo2.max.test.vo2_max_visit1 | -0.468 | 8.37e-05 |
| t66-lung,prot-ph | PC2 | terminal.weight.mg | 0.654 | 4.05e-10 |
| t66-lung,prot-pr | PC2 | registration.sex | -0.683 | 6.66e-08 |
| t66-lung,prot-pr | PC2 | terminal.weight.mg | -0.611 | 1.05e-07 |
| t66-lung,prot-pr | PC4 | registration.sex | -0.601 | 8.39e-07 |
| t66-lung,prot-pr | PC4 | training.staffid | 0.575 | 4.72e-07 |
| t66-lung,prot-pr | PC4 | terminal.weight.mg | -0.494 | 7.34e-06 |
| t68-liver,prot-ph | PC1 | registration.sex | -0.864 | 4.26e-31 |
| t68-liver,prot-ph | PC1 | training.staffid | 0.565 | 2.57e-05 |
| t68-liver,prot-ph | PC1 | vo2.max.test.vo2_max_visit1 | 0.575 | 1.76e-07 |
| t68-liver,prot-ph | PC1 | terminal.weight.mg | -0.714 | 3.91e-22 |
| t68-liver,prot-pr | PC1 | registration.sex | -0.864 | 1.58e-34 |
| t68-liver,prot-pr | PC1 | vo2.max.test.vo2_max_visit1 | 0.588 | 1.57e-07 |
| t68-liver,prot-pr | PC1 | terminal.weight.mg | -0.735 | 8.31e-24 |
| t68-liver,prot-ac | PC1 | registration.sex | -0.861 | 2.37e-25 |
| t68-liver,prot-ac | PC1 | training.staffid | 0.583 | 3.77e-05 |
| t68-liver,prot-ac | PC1 | vo2.max.test.vo2_max_visit1 | 0.540 | 8.38e-06 |
| t68-liver,prot-ac | PC1 | terminal.weight.mg | -0.677 | 1.21e-18 |
| t68-liver,prot-ac | PC3 | training.staffid | -0.512 | 4.85e-05 |
| t68-liver,prot-ub | PC1 | registration.sex | -0.862 | 8.08e-24 |
| t68-liver,prot-ub | PC1 | training.staffid | 0.599 | 9.65e-07 |
| t68-liver,prot-ub | PC1 | vo2.max.test.vo2_max_visit1 | 0.602 | 1.81e-07 |
| t68-liver,prot-ub | PC1 | terminal.weight.mg | -0.733 | 3.59e-19 |
| t70-white-adipose,prot-ph | PC1 | registration.sex | -0.866 | 1.61e-34 |
| t70-white-adipose,prot-ph | PC1 | training.staffid | 0.476 | 5.32e-05 |
| t70-white-adipose,prot-ph | PC1 | vo2.max.test.vo2_max_visit1 | 0.609 | 1.62e-07 |
| t70-white-adipose,prot-ph | PC1 | terminal.weight.mg | -0.754 | 8.79e-25 |
| t70-white-adipose,prot-pr | PC1 | registration.sex | -0.866 | 2.54e-34 |
| t70-white-adipose,prot-pr | PC1 | training.staffid | 0.539 | 3.19e-05 |
| t70-white-adipose,prot-pr | PC1 | vo2.max.test.vo2_max_visit1 | 0.610 | 9.01e-08 |
| t70-white-adipose,prot-pr | PC1 | terminal.weight.mg | -0.763 | 1.90e-24 |
| t70-white-adipose,prot-pr | PC5 | is_control | -0.599 | 3.04e-07 |
kable(metadata_variable_assoc_report,longtable=TRUE,
caption = "Correlations between metadata variables")%>%
kable_styling(latex_options = c("repeat_header"),font_size = 8)3.4 PCA outliers
In this analysis, outliers are flagged by examining the boxplot of each PC, extending its whiskers to three times the inter-quantile range away from the boxplot. Samples outside this range are then flagged.
Note that samples are flagged using an automatic analysis of the principal components. Such analyses may flag samples because of true biological effects and thus further examination is required before determining if flagged samples represent low quality samples. Moreover, note that we plot outliers for the rii (raw intensitied) data as well. Typically, outliers from such datasets will not appear in the bettern normalized ratio data.
pca_outliers_report = c()
for(dataset_name in names(proteomics_data)){
# Get the projected PCs
curr_pcax = proteomics_data[[dataset_name]][["norm_filtered"]][["pca_b"]][[1]]
# Univariate: use IQRs
pca_outliers = c()
for(j in 1:num_pcs_for_outlier_analysis){
#Identify outlier values
outlier_values <- boxplot.stats(curr_pcax[,j],coef = 3)$out
outlier_id <- curr_pcax[,j] %in% outlier_values
#Extract outlier names
outlier_values <- curr_pcax[outlier_id,j]
names(outlier_values) <- rownames(curr_pcax)[outlier_id]
for(outlier in names(outlier_values)){
pca_outliers_report = rbind(pca_outliers_report,
c(dataset_name,paste("PC",j,sep=""),outlier,
format(outlier_values[outlier],digits=5))
)
if(!is.element(outlier,names(pca_outliers))){
pca_outliers[outlier] = outlier_values[outlier]
}
}
}
# Plot the outliers
if(length(pca_outliers)>0){
df = data.frame(curr_pcax,
outliers = rownames(curr_pcax) %in% names(pca_outliers))
col = rep("black",nrow(df))
col[df$outliers] = "green"
plot(df$PC1,df$PC2,pch = as.numeric(df$outliers),col=col,lwd=2,cex=1,
main = paste(dataset_name,"flagged outliers"),
xlab = "PC1",ylab="PC2")
plot(df$PC3,df$PC4,pch = as.numeric(df$outliers),col=col,lwd=2,cex=1,
main = paste(dataset_name,"flagged outliers"),
xlab = "PC3",ylab="PC4")
}
}
if(!is.null(dim(pca_outliers_report))){
colnames(pca_outliers_report) = c("dataset","PC","sample","score")
}3.5 Sex prediction accuracy
Here we map features to their chromosomes and then use the X and Y chromosomes to train a classifier for predicting sex. Note that unlike the sequencing platforms we use the molecular features from the data, which requires running a regularized learning algorithm. We use linear SVM as it seems to produce reasonable results.
entrez2chr = as.list(org.Rn.egCHR)
symbol2entrez = as.list(org.Rn.egSYMBOL2EG)
sex_pred_err_rates = c()
sex_check_outliers = c()
for(dataset_name in names(proteomics_data)){
curr_data = proteomics_data[[dataset_name]][["norm_filtered"]][[2]]
curr_annot = proteomics_data[[dataset_name]]$row_annot
# get the correct row annotation (was filtered by NAs above)
feature_inds = proteomics_data[[dataset_name]][["norm_filtered"]]$feature_inds
curr_annot = curr_annot[feature_inds,]
# map to entrez ids
curr_entrez = symbol2entrez[curr_annot[,"gene_symbol"]]
# take rows with a unique entrez id
single_entrez = sapply(curr_entrez,length)==1
curr_data = curr_data[single_entrez,]
curr_annot = curr_annot[single_entrez,]
curr_entrez = unlist(curr_entrez[single_entrez])
# map to chromosomes
curr_chrs = entrez2chr[curr_entrez]
# limit the data to X and Y chromosomes
chrr_y_genes = sapply(curr_chrs,function(x)any(x=="Y" | x=="X"))
chrr_y_genes[is.na(chrr_y_genes)] = F
if(sum(chrr_y_genes,na.rm=TRUE)<2){next}
sex = pheno_data$viallabel_data[colnames(curr_data),"registration.sex"]
df = data.frame(sex=as.factor(sex),t(curr_data[chrr_y_genes,]))
df = df[,colSums(is.na(df))==0]
# run an svm classifier
train_control <- caret::trainControl(method = "cv", number = nrow(df),
savePredictions = TRUE)
# train the model on training set
model <- caret::train(sex ~ .,data = df,
trControl = train_control,method = "svmLinear2")
# loocv results
cv_res = model$pred
cv_res = cv_res[cv_res[,"cost"]==0.5,]
cv_errors = cv_res[,1] != cv_res[,2]
acc = 1 - (sum(cv_errors)/nrow(cv_res))
print(paste(
"In dataset",dataset_name,
"sex prediction using X,Y chromosomes, cross-validation accuracy is:",
round(acc,digits = 3)))
sex_pred_err_rates = rbind(sex_pred_err_rates,
c(dataset_name,mean(cv_errors))
)
err_samples = rownames(df)[cv_errors]
for(samp in err_samples){
sex_check_outliers = rbind(
sex_check_outliers,c(dataset_name,samp))
}
}## Warning in nominalTrainWorkflow(x = x, y = y, wts = weights, info = trainInfo, :
## There were missing values in resampled performance measures.## [1] "In dataset t53-cortex,prot-ph sex prediction using X,Y chromosomes, cross-validation accuracy is: 0.8"## Warning in nominalTrainWorkflow(x = x, y = y, wts = weights, info = trainInfo, :
## There were missing values in resampled performance measures.## [1] "In dataset t53-cortex,prot-pr sex prediction using X,Y chromosomes, cross-validation accuracy is: 1"## Warning in nominalTrainWorkflow(x = x, y = y, wts = weights, info = trainInfo, :
## There were missing values in resampled performance measures.## [1] "In dataset t55-gastrocnemius,prot-ph sex prediction using X,Y chromosomes, cross-validation accuracy is: 0.9"## Warning in nominalTrainWorkflow(x = x, y = y, wts = weights, info = trainInfo, :
## There were missing values in resampled performance measures.## [1] "In dataset t55-gastrocnemius,prot-pr sex prediction using X,Y chromosomes, cross-validation accuracy is: 0.983"## Warning in nominalTrainWorkflow(x = x, y = y, wts = weights, info = trainInfo, :
## There were missing values in resampled performance measures.## [1] "In dataset t58-heart,prot-ph sex prediction using X,Y chromosomes, cross-validation accuracy is: 1"## Warning in nominalTrainWorkflow(x = x, y = y, wts = weights, info = trainInfo, :
## There were missing values in resampled performance measures.## [1] "In dataset t58-heart,prot-pr sex prediction using X,Y chromosomes, cross-validation accuracy is: 1"## Warning in nominalTrainWorkflow(x = x, y = y, wts = weights, info = trainInfo, :
## There were missing values in resampled performance measures.## [1] "In dataset t58-heart,prot-ac sex prediction using X,Y chromosomes, cross-validation accuracy is: 0.926"## Warning in nominalTrainWorkflow(x = x, y = y, wts = weights, info = trainInfo, :
## There were missing values in resampled performance measures.## [1] "In dataset t58-heart,prot-ub sex prediction using X,Y chromosomes, cross-validation accuracy is: 0.933"## Warning in nominalTrainWorkflow(x = x, y = y, wts = weights, info = trainInfo, :
## There were missing values in resampled performance measures.## [1] "In dataset t59-kidney,prot-ph sex prediction using X,Y chromosomes, cross-validation accuracy is: 1"## Warning in nominalTrainWorkflow(x = x, y = y, wts = weights, info = trainInfo, :
## There were missing values in resampled performance measures.## [1] "In dataset t59-kidney,prot-pr sex prediction using X,Y chromosomes, cross-validation accuracy is: 1"## Warning in nominalTrainWorkflow(x = x, y = y, wts = weights, info = trainInfo, :
## There were missing values in resampled performance measures.## [1] "In dataset t66-lung,prot-ph sex prediction using X,Y chromosomes, cross-validation accuracy is: 0.967"## Warning in nominalTrainWorkflow(x = x, y = y, wts = weights, info = trainInfo, :
## There were missing values in resampled performance measures.## [1] "In dataset t66-lung,prot-pr sex prediction using X,Y chromosomes, cross-validation accuracy is: 1"## Warning in nominalTrainWorkflow(x = x, y = y, wts = weights, info = trainInfo, :
## There were missing values in resampled performance measures.## [1] "In dataset t68-liver,prot-ph sex prediction using X,Y chromosomes, cross-validation accuracy is: 0.983"## Warning in nominalTrainWorkflow(x = x, y = y, wts = weights, info = trainInfo, :
## There were missing values in resampled performance measures.## [1] "In dataset t68-liver,prot-pr sex prediction using X,Y chromosomes, cross-validation accuracy is: 0.983"## Warning in nominalTrainWorkflow(x = x, y = y, wts = weights, info = trainInfo, :
## There were missing values in resampled performance measures.## [1] "In dataset t68-liver,prot-ac sex prediction using X,Y chromosomes, cross-validation accuracy is: 0.981"## Warning in nominalTrainWorkflow(x = x, y = y, wts = weights, info = trainInfo, :
## There were missing values in resampled performance measures.## [1] "In dataset t68-liver,prot-ub sex prediction using X,Y chromosomes, cross-validation accuracy is: 1"## Warning in nominalTrainWorkflow(x = x, y = y, wts = weights, info = trainInfo, :
## There were missing values in resampled performance measures.## [1] "In dataset t70-white-adipose,prot-ph sex prediction using X,Y chromosomes, cross-validation accuracy is: 0.983"## Warning in nominalTrainWorkflow(x = x, y = y, wts = weights, info = trainInfo, :
## There were missing values in resampled performance measures.## [1] "In dataset t70-white-adipose,prot-pr sex prediction using X,Y chromosomes, cross-validation accuracy is: 1"if(! is.null(dim(sex_check_outliers))){
colnames(sex_check_outliers) = c("dataset","sample")
}3.6 Summary of PCA and sex prediction outliers
sample2dataset = c()
for(dataset_name in names(proteomics_data)){
sample2dataset[colnames(proteomics_data[[dataset_name]]$sample_data)] = dataset_name
}
# add the dataset name to mop outliers
all_flagged = NULL
if(length(pca_outliers_report)>0){
all_flagged = union(all_flagged,pca_outliers_report[,"sample"])
}
if(length(sex_check_outliers)>0){
all_flagged = union(all_flagged,sex_check_outliers[,2])
}
flagged_sample_report = c()
for(samp in all_flagged){
samp_dataset = sample2dataset[samp]
samp_metric = NULL
if(!is.null(dim(sex_check_outliers)) &&
samp %in% sex_check_outliers[,"sample"]){
samp_metric = c(samp_metric,"Sex-flagged")
}
if(!is.null(dim(pca_outliers_report)) &&
samp %in% pca_outliers_report[,"sample"]){
curr_pcs = pca_outliers_report[pca_outliers_report[,"sample"]==samp,"PC"]
samp_metric = c(samp_metric,curr_pcs)
}
flagged_sample_report = rbind(flagged_sample_report,
c(samp,samp_dataset,paste(samp_metric,collapse=","))
)
}
colnames(flagged_sample_report) = c("Viallabel","Dataset","Methods")
kable(flagged_sample_report,longtable=TRUE,
caption = "Outliers detected by tissue PCA data")%>%
kable_styling(font_size = 8,latex_options = c("hold_position", "repeat_header"))| Viallabel | Dataset | Methods |
|---|---|---|
| 90225015305 | t53-cortex,prot-pr | Sex-flagged,PC2,PC3,PC2 |
| 90245015305 | t53-cortex,prot-pr | PC2,PC3 |
| 90421015305 | t53-cortex,prot-pr | PC2,PC3,PC2 |
| 90223015507 | t55-gastrocnemius,prot-pr | Sex-flagged,PC2 |
| 90237015805 | t58-heart,prot-ub | PC3 |
| 90245015805 | t58-heart,prot-ub | PC3 |
| 90432015802 | t58-heart,prot-ub | PC3 |
| 90444015805 | t58-heart,prot-ub | Sex-flagged,PC3 |
| 90559015805 | t58-heart,prot-ub | Sex-flagged,PC2 |
| 90283015305 | t53-cortex,prot-pr | Sex-flagged |
| 90406015305 | t53-cortex,prot-pr | Sex-flagged |
| 90280015305 | t53-cortex,prot-pr | Sex-flagged |
| 90581015305 | t53-cortex,prot-pr | Sex-flagged |
| 90410015305 | t53-cortex,prot-pr | Sex-flagged |
| 90222015305 | t53-cortex,prot-pr | Sex-flagged |
| 90248015305 | t53-cortex,prot-pr | Sex-flagged |
| 90258015305 | t53-cortex,prot-pr | Sex-flagged |
| 90559015305 | t53-cortex,prot-pr | Sex-flagged |
| 90564015305 | t53-cortex,prot-pr | Sex-flagged |
| 90439015305 | t53-cortex,prot-pr | Sex-flagged |
| 90406015507 | t55-gastrocnemius,prot-pr | Sex-flagged |
| 90231015503 | t55-gastrocnemius,prot-pr | Sex-flagged |
| 90259015507 | t55-gastrocnemius,prot-pr | Sex-flagged |
| 90283015507 | t55-gastrocnemius,prot-pr | Sex-flagged |
| 90252015507 | t55-gastrocnemius,prot-pr | Sex-flagged |
| 90258015805 | t58-heart,prot-ub | Sex-flagged |
| 90239015805 | t58-heart,prot-ub | Sex-flagged |
| 90289015805 | t58-heart,prot-ub | Sex-flagged |
| 90252015805 | t58-heart,prot-ub | Sex-flagged |
| 90234015802 | t58-heart,prot-ub | Sex-flagged |
| 90217016609 | t66-lung,prot-pr | Sex-flagged |
| 90255016609 | t66-lung,prot-pr | Sex-flagged |
| 90560016806 | t68-liver,prot-ub | Sex-flagged |
| 90430017006 | t70-white-adipose,prot-pr | Sex-flagged |
4. Remove outlier samples
Specify samples that will be removed from differential enrichment results and from the normalized outputs produced by this notebook.
#Identify samples that need to be removed as a named list for each ome
VIALS_TO_REMOVE <- list(
#Samples removed from global proteome data
"t68-liver,prot-pr" = c("90560016806"),
"t55-gastrocnemius,prot-pr" = c("90252015507","90245015507","90223015507"),
#Samples removed from global phosphoproteome data
"t68-liver,prot-ph" = c("90560016806"),
"t55-gastrocnemius,prot-ph" = c("90252015507","90245015507","90223015507"),
#Samples removed from global acetylome data
"t68-liver,prot-ac" = c("90560016806"),
#Samples removed from global ubiquitylome data
"t58-heart,prot-ub" = c("90559015805")
)This code removes the outlier samples from downstream differential analysis and output tables
#Loop through each dataset to remove outlier samples
for(dataset in names(VIALS_TO_REMOVE)){
#Remove sample from the original sample matrix
proteomics_data[[dataset]]$sample_data <-
proteomics_data[[dataset]]$sample_data %>%
dplyr::select(-all_of(VIALS_TO_REMOVE[[dataset]]))
#Remove samples from the metadata matrix
proteomics_data[[dataset]]$sample_meta <-
proteomics_data[[dataset]]$sample_meta %>%
filter(!(vial_label %in% VIALS_TO_REMOVE[[dataset]]))
# Loop through the various normalized datasets and remove the outlier sample
for(i in names(proteomics_data[[dataset]]$norm_filtered)){
if (!(i %in% c("feature_inds","pca_b","pca_w_b"))){
to_keep <- !(colnames(proteomics_data[[dataset]]$norm_filtered[[i]]) %in% VIALS_TO_REMOVE[[dataset]])
proteomics_data[[dataset]]$norm_filtered[[i]] <-
proteomics_data[[dataset]]$norm_filtered[[i]][,to_keep]
}
}
}5. Remove contaminant proteins
This chunk remove contaminant proteins routinely detected during MS analysis. Contaminants are removed at this step and not earlier as they may be useful for the PTM-Protein correction step and for quality control
for(dataset_name in names(proteomics_data)){
#Identify contaminant features
contaminant_ids <- proteomics_data[[dataset_name]]$row_annot %>%
as.data.frame %>% filter(is_contaminant == TRUE) %>% rownames
#Update feature index
proteomics_data[[dataset_name]]$norm_filtered$feature_inds[contaminant_ids] <- F
# Loop through the various normalized datasets and remove contaminant features
for(i in names(proteomics_data[[dataset]]$norm_filtered)){
if (!(i %in% c("feature_inds","pca_b","pca_w_b"))){
to_keep <- setdiff(rownames(proteomics_data[[dataset_name]]$norm_filtered[[i]]),
contaminant_ids)
proteomics_data[[dataset_name]]$norm_filtered[[i]] <-
proteomics_data[[dataset_name]]$norm_filtered[[i]][to_keep,]
}
}
}6. Differential analysis
Differential analysis is performed using two approaches. Section 6.1 uses a model that combines both animal sexes, while section 6.2 fits a model on samples from each sex. Only results from section 6.2 are saved and uploaded to the Google bucket.
Future implementations can modify Section 6.1 to account for sex-specific residuals. For example, by using generalized least squares.
6.1 DEA without splitting sexes
Here, we use a combined model for both sexes.
First, perform DEA on non-corrected data
# a helper function for getting SEs per estimate in the timewise
# data analysis
limma_res_extract_se<-function(limma_res,e_fit,
effect_col="logFC",t_col="t",colname=NULL){
# method 1
effects = limma_res[[effect_col]]
ts = limma_res[[t_col]]
ses1 = effects/ts
if(is.null(colname)){
return(ses1)
}
# method 2
ses2 = sqrt(e_fit$s2.post) * e_fit$stdev.unscaled
ts1 = effects/ses1
ts2 = effects/ses2[,colname]
if(max(abs(ts-ts1),na.rm = T) > 1e-10 ||
max(abs(ts-ts2),na.rm = T) > 1e-10 ||
max(abs(ts1-ts2),na.rm = T)>1e-10){
print("Warning: t-stats from computed ses are incompatible with limma's output")
}
return(ses2[,colname])
}
ftest_res = c() # keeps all ftest results
sex_ftest_res = c() # keeps the sex-specific f-tests
dataset2num_discoveries = c() # count stats
tpDA = c() # keep the timewise results
for(dataset_name in names(proteomics_data)){
#Extract current dataset
x = proteomics_data[[dataset_name]][["norm_filtered"]][[1]]
#Extract sample annotations
x_annot = proteomics_data[[dataset_name]]$row_annot
feature_inds = proteomics_data[[dataset_name]][["norm_filtered"]]$feature_inds
x_annot = x_annot[feature_inds,]
covs = data.frame(
sex = pheno_data$viallabel_data[colnames(x),"registration.sex"],
timepoint = pheno_data$viallabel_data[colnames(x),"timepoint"],
is_control = pheno_data$viallabel_data[colnames(x),"is_control"]
)
covs$sex[covs$sex=="2"] = "M"; covs$sex[covs$sex=="1"]="F"
rownames(covs) = colnames(x)
#Specify the model
full_model_str = "~1+sex+group+sex:group"
# F-test analysis - training-dea table
# Fit a single limma model for F-test.
# Perform F-test using the exercise groups and group:sex interactions
# see https://support.bioconductor.org/p/12119/ and the limma guide
group=factor(paste(covs$timepoint,covs$is_control,sep="_"),
levels=c("8_1","1_0","2_0", "4_0", "8_0"))
covs$group=group
des = model.matrix(as.formula(full_model_str),data=covs)
is_group_variable = grepl("group",colnames(des))
colnames(des) = gsub("group","",colnames(des))
colnames(des)[1] = "Intercept"
# fit the models
limma_model1 = lmFit(x,as.matrix(des))
eb_Ftest = eBayes(limma_model1)
# Extract the results
res <- topTable(eb_Ftest,coef = which(is_group_variable),
number = Inf, sort.by = "none")
num_sig1 = sum(p.adjust(res$P.Value,method="fdr")<0.1,na.rm=TRUE)
curr_f_test_res = data.frame(
feature_ID = rownames(x),
assay = strsplit(dataset_name,split=",")[[1]][2],
tissue = strsplit(dataset_name,split=",")[[1]][1],
p_value = res$P.Value,
adj_p_value = p.adjust(res$P.Value,method="fdr"),
fscore = res$`F`,
numNAs = rowSums(is.na(x)),
full_model = full_model_str,
reduced_model = "~1+sex"
)
ftest_res = rbind(ftest_res,curr_f_test_res)
# F-test analysis - training-sex-biased table
# Perform F-test using the group:sex interactions
inter_terms = grepl(":",colnames(des))
res_sex <- topTable(eb_Ftest,coef = which(inter_terms),
number = Inf, sort.by = "none")
num_sig_sex = sum(p.adjust(res_sex$P.Value,method="fdr")<0.1,na.rm=TRUE)
curr_sex_f_test_res = data.frame(
feature_ID = rownames(x),
assay = strsplit(dataset_name,split=",")[[1]][2],
tissue = strsplit(dataset_name,split=",")[[1]][1],
p_value = res_sex$P.Value,
adj_p_value = p.adjust(res_sex$P.Value,method="fdr"),
fscore = res_sex$`F`,
numNAs = rowSums(is.na(x)),
full_model = "~1+sex+group+sex:group",
reduced_model = "~1+sex+group"
)
sex_ftest_res = rbind(sex_ftest_res,curr_sex_f_test_res)
# T-test analysis - timewise-dea table
# Perform T-test using the exercise groups and group:sex interactions
group=factor(paste(covs$sex,covs$timepoint,covs$is_control,sep="_"))
des = model.matrix(~0+group)
colnames(des) = gsub("group","",colnames(des))
C_ttests = makeContrasts(
M_1_0 - M_8_1,M_2_0 - M_8_1,M_4_0 - M_8_1,M_8_0 - M_8_1,
F_1_0 - F_8_1,F_2_0 - F_8_1,F_4_0 - F_8_1,F_8_0 - F_8_1,
levels = des
)
# fit the new model
limma_model2 = lmFit(x,as.matrix(des))
lmfit.cont <- contrasts.fit(limma_model2, C_ttests)
lmfit.cont.ebayes <- eBayes(lmfit.cont)
# count the results at 10% FDR
t_ps = c(lmfit.cont.ebayes$p.value)
suppressWarnings({pthr = max(-1,max(t_ps[p.adjust(t_ps,method="BY")<0.1]))})
num_sig2 = sum(apply(lmfit.cont.ebayes$p.value<pthr,1,any),na.rm = T)
suppressWarnings({pthr = max(-1,max(t_ps[p.adjust(t_ps,method="fdr")<0.1]))})
num_sig3 = sum(apply(lmfit.cont.ebayes$p.value<pthr,1,any),na.rm = T)
v_counts = c(dataset=dataset_name,ftest=num_sig1,
sex_ftest=num_sig_sex,ttests_BY=num_sig2,ttests_BH=num_sig3)
dataset2num_discoveries = rbind(dataset2num_discoveries,v_counts)
# extract contrast info
for(col in colnames(lmfit.cont.ebayes$t)){
curr_sex = strsplit(col,split="_")[[1]][1]
sex_str = "male"
if(curr_sex=="F"){sex_str="female"}
curr_tp = strsplit(col,split="_")[[1]][2]
samps = rownames(covs)[covs$sex==curr_sex &
(covs$timepoint==curr_tp | covs$is_control==1)]
if(any(table(covs[samps,"is_case"])<2)){
print(paste("in dataset:",dataset,"skipping time point",col,"of sex",sex))
print(paste("reason: there is a group with less than 2 samples"))
next
}
limma_res = topTable(lmfit.cont.ebayes,
coef = col,number = nrow(x),sort.by = "none")
curr_res = data.frame(
feature_ID = rownames(x),
assay = strsplit(dataset_name,split=",")[[1]][2],
tissue = strsplit(dataset_name,split=",")[[1]][1],
sex = sex_str,
logFC_se = limma_res_extract_se(limma_res,lmfit.cont.ebayes),
logFC = limma_res$logFC,
tscore = limma_res$t,
covariates = NA,
comparison_group = paste0(curr_tp,"w"),
p_value = limma_res$P.Value,
adj_p_value = limma_res$adj.P.Val
)
# add group info
case_samps = samps[covs[samps,"is_case"]=="1"]
curr_res$comparison_average_intensity = apply(x[,case_samps],1,mean)
control_samps = samps[covs[samps,"is_case"]!="1"]
curr_res$reference_average_intensity = apply(x[,control_samps],1,mean)
# add NA counts
curr_res$numNAs = rowSums(is.na(x[,samps]))
# add the results
tpDA = rbind(tpDA,curr_res)
}
}Perform DEA on PTM protein-corrected data.
#Perform DEA only on PTM data
ptm_datasets <- grep("prot-pr",names(proteomics_data),value = T,invert = T)
for(dataset_name in ptm_datasets){
x = proteomics_data[[dataset_name]][["norm_filtered"]][["norm_data_b_protein_corrected"]]
#Extract sample annotations
x_annot = proteomics_data[[dataset_name]]$row_annot
feature_inds = proteomics_data[[dataset_name]][["norm_filtered"]]$feature_inds
x_annot = x_annot[feature_inds,]
covs = data.frame(
sex = pheno_data$viallabel_data[colnames(x),"registration.sex"],
timepoint = pheno_data$viallabel_data[colnames(x),"timepoint"],
is_control = pheno_data$viallabel_data[colnames(x),"is_control"]
)
covs$sex[covs$sex=="2"] = "M";covs$sex[covs$sex=="1"]="F"
rownames(covs) = colnames(x)
#Specify the model
full_model_str = "~1+sex+group+sex:group"
# F-test analysis - training-dea table
# Fit a single limma model for F-test.
# Perform F-test using the exercise groups and group:sex interactions
# see https://support.bioconductor.org/p/12119/ and the limma guide
group=factor(paste(covs$timepoint,covs$is_control,sep="_"),
levels=c("8_1","1_0","2_0", "4_0", "8_0"))
covs$group=group
des = model.matrix(as.formula(full_model_str),data=covs)
is_group_variable = grepl("group",colnames(des))
colnames(des) = gsub("group","",colnames(des))
colnames(des)[1] = "Intercept"
# fit the models
limma_model1 = lmFit(x,as.matrix(des))
eb_Ftest = eBayes(limma_model1)
# Extract the results
res <- topTable(eb_Ftest,coef = which(is_group_variable),
number = Inf, sort.by = "none")
num_sig1 = sum(p.adjust(res$P.Value,method="fdr")<0.1,na.rm=TRUE)
curr_f_test_res = data.frame(
feature_ID = rownames(x),
assay = paste0(strsplit(dataset_name,split=",")[[1]][2],"-protein-corrected"),
tissue = strsplit(dataset_name,split=",")[[1]][1],
p_value = res$P.Value,
adj_p_value = p.adjust(res$P.Value,method="fdr"),
fscore = res$`F`,
numNAs = rowSums(is.na(x)),
full_model = full_model_str,
reduced_model = "~1+sex"
)
ftest_res = rbind(ftest_res,curr_f_test_res)
# F-test analysis - training-sex-biased table
# Perform F-test using the group:sex interactions
inter_terms = grepl(":",colnames(des))
res_sex <- topTable(eb_Ftest,coef = which(inter_terms),
number = Inf, sort.by = "none")
num_sig_sex = sum(p.adjust(res_sex$P.Value,method="fdr")<0.1,na.rm=TRUE)
curr_sex_f_test_res = data.frame(
feature_ID = rownames(x),
assay = paste0(strsplit(dataset_name,split=",")[[1]][2],"-protein-corrected"),
tissue = strsplit(dataset_name,split=",")[[1]][1],
p_value = res_sex$P.Value,
adj_p_value = p.adjust(res_sex$P.Value,method="fdr"),
fscore = res_sex$`F`,
numNAs = rowSums(is.na(x)),
full_model = "~1+sex+group+sex:group",
reduced_model = "~1+sex+group"
)
sex_ftest_res = rbind(sex_ftest_res,curr_sex_f_test_res)
# T-test analysis - timewise-dea table
# Perform T-test using the exercise groups and group:sex interactions
group=factor(paste(covs$sex,covs$timepoint,covs$is_control,sep="_"))
des = model.matrix(~0+group)
colnames(des) = gsub("group","",colnames(des))
C_ttests = makeContrasts(
M_1_0 - M_8_1,M_2_0 - M_8_1,M_4_0 - M_8_1,M_8_0 - M_8_1,
F_1_0 - F_8_1,F_2_0 - F_8_1,F_4_0 - F_8_1,F_8_0 - F_8_1,
levels = des
)
# fit the new model
limma_model2 = lmFit(x,as.matrix(des))
lmfit.cont <- contrasts.fit(limma_model2, C_ttests)
lmfit.cont.ebayes <- eBayes(lmfit.cont)
# count the results at 10% FDR
t_ps = c(lmfit.cont.ebayes$p.value)
suppressWarnings({pthr = max(-1,max(t_ps[p.adjust(t_ps,method="BY")<0.1]))})
num_sig2 = sum(apply(lmfit.cont.ebayes$p.value<pthr,1,any),na.rm = T)
suppressWarnings({pthr = max(-1,max(t_ps[p.adjust(t_ps,method="fdr")<0.1]))})
num_sig3 = sum(apply(lmfit.cont.ebayes$p.value<pthr,1,any),na.rm = T)
v_counts = c(dataset=dataset_name,ftest=num_sig1,
sex_ftest=num_sig_sex,ttests_BY=num_sig2,ttests_BH=num_sig3)
dataset2num_discoveries = rbind(dataset2num_discoveries,v_counts)
# extract contrast info
for(col in colnames(lmfit.cont.ebayes$t)){
curr_sex = strsplit(col,split="_")[[1]][1]
sex_str = "male"
if(curr_sex=="F"){sex_str="female"}
curr_tp = strsplit(col,split="_")[[1]][2]
samps = rownames(covs)[covs$sex==curr_sex &
(covs$timepoint==curr_tp | covs$is_control==1)]
if(any(table(covs[samps,"is_case"])<2)){
print(paste("in dataset:",dataset,"skipping time point",col,"of sex",sex))
print(paste("reason: there is a group with less than 2 samples"))
next
}
limma_res = topTable(lmfit.cont.ebayes,
coef = col,number = nrow(x),sort.by = "none")
curr_res = data.frame(
feature_ID = rownames(x),
assay = paste0(strsplit(dataset_name,split=",")[[1]][2],"-protein-corrected"),
tissue = strsplit(dataset_name,split=",")[[1]][1],
sex = sex_str,
logFC_se = limma_res_extract_se(limma_res,lmfit.cont.ebayes),
logFC = limma_res$logFC,
tscore = limma_res$t,
covariates = NA,
comparison_group = paste0(curr_tp,"w"),
p_value = limma_res$P.Value,
adj_p_value = limma_res$adj.P.Val
)
# add group info
case_samps = samps[covs[samps,"is_case"]=="1"]
curr_res$comparison_average_intensity = apply(x[,case_samps],1,mean,na.rm=TRUE)
control_samps = samps[covs[samps,"is_case"]!="1"]
curr_res$reference_average_intensity = apply(x[,control_samps],1,mean,na.rm=TRUE)
# add NA counts
curr_res$numNAs = rowSums(is.na(x[,samps]))
# add the results
tpDA = rbind(tpDA,curr_res)
}
}6.2 DEA separating sexes
Here, we perform DEA by fitting models to only Male or only Female samples.
First run on non-corrected data.
ftest_res_split_sex = c() # keeps all ftest results
tpDA_split_sex = c() # keep the timewise results
for(dataset_name in names(proteomics_data)){
#Extract current dataset and metadata
x = proteomics_data[[dataset_name]][["norm_filtered"]][[1]]
x_annot = proteomics_data[[dataset_name]]$row_annot
feature_inds = proteomics_data[[dataset_name]][["norm_filtered"]]$feature_inds
x_annot = x_annot[feature_inds,]
#Specify covariates
covs = data.frame(
sex = pheno_data$viallabel_data[colnames(x),"registration.sex"],
timepoint = pheno_data$viallabel_data[colnames(x),"timepoint"],
is_control = pheno_data$viallabel_data[colnames(x),"is_control"]
)
covs$sex[covs$sex=="2"] = "M";covs$sex[covs$sex=="1"]="F"
rownames(covs) = colnames(x)
#Variables that save sex-specific results
sex_res = list() #F-test result
sex_ttest_res = list() #T-test result
for(SEX in c('M','F')){
curr_meta = covs %>% filter(sex == SEX) %>% mutate(group = paste(timepoint,is_control,sep="_"))
curr_counts = x[,rownames(curr_meta)]
###################################################################
# F-test analysis - training-dea table
#Extract treatment covariate
tr <- factor(paste(curr_meta$timepoint,curr_meta$is_control,sep="_"),
levels=c("8_1","1_0","2_0","4_0","8_0"))
#Generate the experimental model
design <- model.matrix(~ 1+tr)
fit <- lmFit(curr_counts, design)
fit.eb <- eBayes(fit)
#Extract results
res = topTable(fit.eb, coef = 2:5, n=nrow(fit.eb))
dt = data.table(tissue=str_split(dataset_name,",")[[1]][1],
assay=str_split(dataset_name,",")[[1]][2],
feature_ID=rownames(res),
fscore=res$`F`,
p_value = res$P.Value,
full_model = "~1+group",
reduced_model = "~1")
sex_res[[SEX]] = dt
###############################################################
# T-test analysis - timewise-dea table
# Perform T-test using the exercise groups
design = model.matrix(~0+tr)
#Set contrasts
cont.matrix = makeContrasts(
tr1_0 - tr8_1, tr2_0 - tr8_1, tr4_0 - tr8_1, tr8_0 - tr8_1,
levels = design
)
colnames(cont.matrix) = c("1w","2w","4w","8w")
# Fit the new model
limma_model2 = lmFit(curr_counts,design)
lmfit.cont <- contrasts.fit(limma_model2, cont.matrix)
lmfit.cont.ebayes <- eBayes(lmfit.cont)
#Extract results for each timepoint
for(curr_tp in colnames(lmfit.cont.ebayes$t)){
#Extract results
limma_res = topTable(lmfit.cont.ebayes,
coef = curr_tp,number = Inf,sort.by = "none")
curr_res = data.frame(
feature_ID = rownames(limma_res),
assay = strsplit(dataset_name,split=",")[[1]][2],
tissue = strsplit(dataset_name,split=",")[[1]][1],
sex = if_else(SEX == "M","male","female"),
logFC_se = limma_res_extract_se(limma_res,lmfit.cont.ebayes),
logFC = limma_res$logFC,
tscore = limma_res$t,
covariates = NA,
comparison_group = curr_tp,
p_value = limma_res$P.Value
)
# Add group average intensities
case_samps = covs %>%
filter(sex == SEX &
timepoint == substr(curr_tp,1,1)&
is_control == 0) %>%
rownames()
curr_res$comparison_average_intensity = apply(x[,case_samps],1,mean,na.rm=TRUE)
control_samps = covs %>%
filter(sex == SEX &
is_control == 1) %>%
rownames()
curr_res$reference_average_intensity = apply(x[,control_samps],1,mean,na.rm=TRUE)
# Add NA counts
curr_res$numNAs = rowSums(is.na(x[,c(case_samps,control_samps)]))
# Add the results
tpDA_split_sex = rbind(tpDA_split_sex,curr_res)
}
}
#Merge F-test results
merged = data.frame(merge(sex_res[['M']], sex_res[['F']],
by=c("tissue","assay","feature_ID","full_model","reduced_model"),
suffixes=c('_male','_female'))) %>%
mutate(p_value_male = replace_na(p_value_male,1),
p_value_female = replace_na(p_value_female,1)) %>%
mutate(p_value = map2_dbl(p_value_male,p_value_female,function(x,y){sumlog(c(x,y))$p}))
ftest_res_split_sex <- rbind(ftest_res_split_sex,merged)
}Next, run on protein-corrected PTM data
#Perform DEA only on PTM data
ptm_datasets <- grep("prot-pr", names(proteomics_data),value = T,invert = T)
for(dataset_name in ptm_datasets){
#Extract current dataset and metadata
x = proteomics_data[[dataset_name]][["norm_filtered"]][["norm_data_b_protein_corrected"]]
x_annot = proteomics_data[[dataset_name]]$row_annot
feature_inds = proteomics_data[[dataset_name]][["norm_filtered"]]$feature_inds
x_annot = x_annot[feature_inds,]
#Specify covariates
covs = data.frame(
sex = pheno_data$viallabel_data[colnames(x),"registration.sex"],
timepoint = pheno_data$viallabel_data[colnames(x),"timepoint"],
is_control = pheno_data$viallabel_data[colnames(x),"is_control"]
)
covs$sex[covs$sex=="2"] = "M";covs$sex[covs$sex=="1"]="F"
rownames(covs) = colnames(x)
#Variables that save sex-specific results
sex_res = list() #F-test result
sex_ttest_res = list() #T-test result
for(SEX in c('M','F')){
curr_meta = covs %>% filter(sex == SEX) %>% mutate(group = paste(timepoint,is_control,sep="_"))
curr_counts = x[,rownames(curr_meta)]
###################################################################
# F-test analysis - training-dea table
#Extract treatment covariate
tr <- factor(paste(curr_meta$timepoint,curr_meta$is_control,sep="_"),
levels=c("8_1","1_0","2_0","4_0","8_0"))
#Generate the experimental model
design <- model.matrix(~ 1+tr)
fit <- lmFit(curr_counts, design)
fit.eb <- eBayes(fit)
#Extract results
res = topTable(fit.eb, coef = 2:5, n=nrow(fit.eb))
dt = data.table(
tissue=str_split(dataset_name,",")[[1]][1],
assay = paste0(strsplit(dataset_name,split=",")[[1]][2],"-protein-corrected"),
feature_ID=rownames(res),
fscore=res$`F`,
p_value = res$P.Value,
full_model = "~1+group",
reduced_model = "~1")
sex_res[[SEX]] = dt
###############################################################
# T-test analysis - timewise-dea table
# Perform T-test using the exercise groups
design = model.matrix(~0+tr)
#Set contrasts
cont.matrix = makeContrasts(
tr1_0 - tr8_1, tr2_0 - tr8_1, tr4_0 - tr8_1, tr8_0 - tr8_1,
levels = design
)
colnames(cont.matrix) = c("1w","2w","4w","8w")
# Fit the new model
limma_model2 = lmFit(curr_counts,design)
lmfit.cont <- contrasts.fit(limma_model2, cont.matrix)
lmfit.cont.ebayes <- eBayes(lmfit.cont)
#Extract results for each timepoint
for(curr_tp in colnames(lmfit.cont.ebayes$t)){
#Extract results
limma_res = topTable(lmfit.cont.ebayes,
coef = curr_tp,
number = Inf,
sort.by = "none")
curr_res = data.frame(
feature_ID = rownames(limma_res),
assay = paste0(strsplit(dataset_name,split=",")[[1]][2],"-protein-corrected"),
tissue = strsplit(dataset_name,split=",")[[1]][1],
sex = if_else(SEX == "M","male","female"),
logFC_se = limma_res_extract_se(limma_res,lmfit.cont.ebayes),
logFC = limma_res$logFC,
tscore = limma_res$t,
covariates = NA,
comparison_group = curr_tp,
p_value = limma_res$P.Value
)
# Add group average intensities
case_samps = covs %>%
filter(sex == SEX &
timepoint == substr(curr_tp,1,1)&
is_control == 0) %>%
rownames()
curr_res$comparison_average_intensity = apply(x[,case_samps],1,mean,na.rm=TRUE)
control_samps = covs %>%
filter(sex == SEX &
is_control == 1) %>%
rownames()
curr_res$reference_average_intensity = apply(x[,control_samps],1,mean,na.rm=TRUE)
# Add NA counts
curr_res$numNAs = rowSums(is.na(x[,c(case_samps,control_samps)]))
# Add the results
tpDA_split_sex = rbind(tpDA_split_sex,curr_res)
}
}
#Merge F-test results
merged = data.frame(merge(sex_res[['M']], sex_res[['F']],
by=c("tissue","assay","feature_ID","full_model","reduced_model"),
suffixes=c('_male','_female'))) %>%
mutate(p_value_male = replace_na(p_value_male,1),
p_value_female = replace_na(p_value_female,1)) %>%
mutate(p_value = map2_dbl(p_value_male,p_value_female,function(x,y){sumlog(c(x,y))$p}))
ftest_res_split_sex <- rbind(ftest_res_split_sex,merged)
}Scatter plots are displayed below to assess the effects of protein correction of PTM data on the DEA results.
ptm_datasets <- grep("prot-pr", names(proteomics_data),value = T,invert = T)
for(dataset_name in ptm_datasets){
cur_tissue <- str_split_fixed(dataset_name,",",2)[1]
cur_assay <- str_split_fixed(dataset_name,",",2)[2]
uncorrected <- ftest_res_split_sex %>%
filter(tissue == cur_tissue &
assay == cur_assay)
corrected <- ftest_res_split_sex %>%
filter(tissue == cur_tissue &
(assay == paste0(cur_assay,"-protein-corrected")))
dat <- inner_join(uncorrected,corrected,
by = c("tissue","feature_ID","full_model","reduced_model"),
suffix = c(".uncorrected",".corrected"))
g <- ggplot(dat, aes(x = -log10(p_value.uncorrected),
y = -log10(p_value.corrected)))+
geom_point(alpha=0.01)+
theme_bw()+
geom_smooth(method="lm",formula = y~x)+
labs(x = "Uncorrected PTM training-dea\n-log10(p-val)",
y = "Protein-corrected PTM training-dea\n-log10(p-val)",
title= dataset_name)
plot(g)
}
6.3 Save DEA results
Save DEA results that split datasets by sex.
tissue_assay_comb = unique(tpDA_split_sex[,c("tissue","assay")])
# timewise tables
for(i in 1:nrow(tissue_assay_comb)){
tissue = tissue_assay_comb[i,1]
assay = tissue_assay_comb[i,2]
currDA = tpDA_split_sex[tpDA_split_sex$tissue == tissue & tpDA_split_sex$assay == assay,]
local_fname = file.path(local_data_dir, paste0("pass1b-06_",tissue,"_",assay,"_timewise-dea_",suffix))
curr_out_bucket = paste0(out_bucket,str_extract(assay,"prot-.."),"/",dea_folder_name,"/")
message("Write Local: ", local_fname)
write.table(currDA,
file=local_fname,
sep="\t",
quote=FALSE,
row.names = FALSE,
col.names = TRUE)
if(upload_to_bucket){
cmd = paste(gsutil_cmd, "-m cp", local_fname, curr_out_bucket)
message("Copy to Bucket: ", cmd)
system(cmd)
}
}## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/pass1b-06_t53-cortex_prot-ph_timewise-dea_20210914.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/pass1b-06_t53-cortex_prot-pr_timewise-dea_20210914.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/pass1b-06_t55-gastrocnemius_prot-ph_timewise-dea_20210914.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/pass1b-06_t55-gastrocnemius_prot-pr_timewise-dea_20210914.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/pass1b-06_t58-heart_prot-ph_timewise-dea_20210914.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/pass1b-06_t58-heart_prot-pr_timewise-dea_20210914.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/pass1b-06_t58-heart_prot-ac_timewise-dea_20210914.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/pass1b-06_t58-heart_prot-ub_timewise-dea_20210914.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/pass1b-06_t59-kidney_prot-ph_timewise-dea_20210914.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/pass1b-06_t59-kidney_prot-pr_timewise-dea_20210914.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/pass1b-06_t66-lung_prot-ph_timewise-dea_20210914.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/pass1b-06_t66-lung_prot-pr_timewise-dea_20210914.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/pass1b-06_t68-liver_prot-ph_timewise-dea_20210914.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/pass1b-06_t68-liver_prot-pr_timewise-dea_20210914.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/pass1b-06_t68-liver_prot-ac_timewise-dea_20210914.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/pass1b-06_t68-liver_prot-ub_timewise-dea_20210914.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/pass1b-06_t70-white-adipose_prot-ph_timewise-dea_20210914.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/pass1b-06_t70-white-adipose_prot-pr_timewise-dea_20210914.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/pass1b-06_t53-cortex_prot-ph-protein-corrected_timewise-dea_20210914.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/pass1b-06_t55-gastrocnemius_prot-ph-protein-corrected_timewise-dea_20210914.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/pass1b-06_t58-heart_prot-ph-protein-corrected_timewise-dea_20210914.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/pass1b-06_t58-heart_prot-ac-protein-corrected_timewise-dea_20210914.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/pass1b-06_t58-heart_prot-ub-protein-corrected_timewise-dea_20210914.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/pass1b-06_t59-kidney_prot-ph-protein-corrected_timewise-dea_20210914.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/pass1b-06_t66-lung_prot-ph-protein-corrected_timewise-dea_20210914.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/pass1b-06_t68-liver_prot-ph-protein-corrected_timewise-dea_20210914.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/pass1b-06_t68-liver_prot-ac-protein-corrected_timewise-dea_20210914.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/pass1b-06_t68-liver_prot-ub-protein-corrected_timewise-dea_20210914.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/pass1b-06_t70-white-adipose_prot-ph-protein-corrected_timewise-dea_20210914.txt# f-test tables
for(i in 1:nrow(tissue_assay_comb)){
tissue = tissue_assay_comb[i,1]
assay = tissue_assay_comb[i,2]
currTable = ftest_res_split_sex[ftest_res_split_sex$tissue == tissue & ftest_res_split_sex$assay==assay,]
local_fname = file.path(local_data_dir, paste0("pass1b-06_",tissue,"_",assay,"_training-dea_",suffix))
curr_out_bucket = paste0(out_bucket,str_extract(assay,"prot-.."),"/",dea_folder_name,"/")
message("Write Local: ", local_fname)
write.table(currTable,
file=local_fname,
sep="\t",
quote=FALSE,
row.names = FALSE,
col.names = TRUE)
if(upload_to_bucket){
cmd = paste(gsutil_cmd,"-m cp",local_fname,curr_out_bucket)
message("\tCopy to Bucket: ", cmd)
system(cmd)
}
}## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/pass1b-06_t53-cortex_prot-ph_training-dea_20210914.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/pass1b-06_t53-cortex_prot-pr_training-dea_20210914.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/pass1b-06_t55-gastrocnemius_prot-ph_training-dea_20210914.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/pass1b-06_t55-gastrocnemius_prot-pr_training-dea_20210914.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/pass1b-06_t58-heart_prot-ph_training-dea_20210914.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/pass1b-06_t58-heart_prot-pr_training-dea_20210914.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/pass1b-06_t58-heart_prot-ac_training-dea_20210914.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/pass1b-06_t58-heart_prot-ub_training-dea_20210914.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/pass1b-06_t59-kidney_prot-ph_training-dea_20210914.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/pass1b-06_t59-kidney_prot-pr_training-dea_20210914.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/pass1b-06_t66-lung_prot-ph_training-dea_20210914.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/pass1b-06_t66-lung_prot-pr_training-dea_20210914.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/pass1b-06_t68-liver_prot-ph_training-dea_20210914.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/pass1b-06_t68-liver_prot-pr_training-dea_20210914.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/pass1b-06_t68-liver_prot-ac_training-dea_20210914.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/pass1b-06_t68-liver_prot-ub_training-dea_20210914.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/pass1b-06_t70-white-adipose_prot-ph_training-dea_20210914.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/pass1b-06_t70-white-adipose_prot-pr_training-dea_20210914.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/pass1b-06_t53-cortex_prot-ph-protein-corrected_training-dea_20210914.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/pass1b-06_t55-gastrocnemius_prot-ph-protein-corrected_training-dea_20210914.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/pass1b-06_t58-heart_prot-ph-protein-corrected_training-dea_20210914.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/pass1b-06_t58-heart_prot-ac-protein-corrected_training-dea_20210914.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/pass1b-06_t58-heart_prot-ub-protein-corrected_training-dea_20210914.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/pass1b-06_t59-kidney_prot-ph-protein-corrected_training-dea_20210914.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/pass1b-06_t66-lung_prot-ph-protein-corrected_training-dea_20210914.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/pass1b-06_t68-liver_prot-ph-protein-corrected_training-dea_20210914.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/pass1b-06_t68-liver_prot-ac-protein-corrected_training-dea_20210914.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/pass1b-06_t68-liver_prot-ub-protein-corrected_training-dea_20210914.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/pass1b-06_t70-white-adipose_prot-ph-protein-corrected_training-dea_20210914.txt7. Save normalized data to output bucket
Save normalized and batch corrected data
for(dataset_name in names(proteomics_data)){
#Normalized and batch corrected data
x = proteomics_data[[dataset_name]][["norm_filtered"]][["norm_data_b"]]
#Normalized, batch corrected, and imputed
x_imp = proteomics_data[[dataset_name]][["norm_filtered"]][["norm_data_b_imp"]]
#Feature annotations
x_annot = proteomics_data[[dataset_name]]$row_annot
# get the correct row annotation (was filtered by NAs above)
feature_inds = proteomics_data[[dataset_name]][["norm_filtered"]]$feature_inds
x_annot = x_annot[feature_inds,]
arr = strsplit(dataset_name,split=",")[[1]]
curr_ome = tolower(arr[2])
curr_t = tolower(arr[1])
out_b = paste0(out_bucket,curr_ome,"/data/")
# add bid and pid
x_bids = pheno_data$viallabel_data[colnames(x),"bid"]
x_pids = pheno_data$viallabel_data[colnames(x),"pid"]
newx = x
for(j in 1:ncol(newx)){newx[,j] = round(as.numeric(newx[,j]),digits=5)}
newx = rbind(x_bids,newx)
newx = rbind(x_pids,newx)
newx = rbind(colnames(x),newx)
colnames(newx) = NULL
newx_imp = x_imp
for(j in 1:ncol(newx_imp)){newx_imp[,j] = round(as.numeric(newx_imp[,j]),digits=5)}
newx_imp = rbind(x_bids,newx_imp)
newx_imp = rbind(x_pids,newx_imp)
newx_imp = rbind(colnames(x_imp),newx_imp)
colnames(newx_imp) = NULL
# a simple sanity check
m = newx[-c(1:3),]
mode(m) = "numeric"
if(!all(abs(m-x)< 1e-05,na.rm=TRUE)){
print("Error in parsing the matrix, breaking")
break
}
# a simple sanity check
m = newx_imp[-c(1:3),]
mode(m) = "numeric"
if(!all(abs(m-x)< 1e-05,na.rm=TRUE)){
print("Error in parsing the matrix, breaking")
break
}
rownames(newx)[1:3] = c("viallabel","pid","bid")
rownames(newx_imp)[1:3] = c("viallabel","pid","bid")
#Save the output to the target bucket
# Median-MAD normalized TMT ratios
local_fname = file.path(local_data_dir,
paste0("motrpac_pass1b-06_",
curr_t,"_",curr_ome,
"_med-mad-normalized-logratio.txt"))
message("Write Local: ", local_fname)
write.table(newx,
file=local_fname,
sep="\t",
quote=FALSE,
row.names = TRUE,
col.names = FALSE)
if(upload_to_bucket){
cmd = paste(gsutil_cmd,"cp",local_fname,out_b)
message("Copy to Bucket: ", cmd)
system(cmd)
}
# Median-MAD normalized and imputed TMT ratios
local_fname_imp = file.path(local_data_dir,
paste0("motrpac_pass1b-06_",
curr_t,"_",curr_ome,
"_med-mad-normalized-imputed-logratio.txt"))
message("Write Local: ", local_fname_imp)
write.table(newx_imp,
file=local_fname_imp,
sep="\t",
quote=FALSE,
row.names = TRUE,
col.names = FALSE)
if(upload_to_bucket){
cmd = paste(gsutil_cmd,"cp",local_fname_imp,out_b)
message("Copy to Bucket: ", cmd)
system(cmd)
}
#Feature annotations
local_fname2 = file.path(local_data_dir,
paste0("motrpac_pass1b-06_",
curr_t,"_",curr_ome,
"_normalized-data-feature-annot.txt"))
x_annot = rbind(colnames(x_annot),x_annot)
colnames(x_annot) = NULL
x_annot = cbind(rownames(x_annot),x_annot)
rownames(x_annot) = NULL
x_annot[,1] = as.character(x_annot[,1])
x_annot[1,1] = "feature_ID"
message("Write Local: ", local_fname2)
write.table(x_annot,
file=local_fname2,
sep="\t",
quote=FALSE,
row.names = FALSE,
col.names = FALSE)
if(upload_to_bucket){
cmd = paste(gsutil_cmd,"cp",local_fname2,out_b)
message("Copy to Bucket: ", cmd)
system(cmd)
}
}## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/motrpac_pass1b-06_t53-cortex_prot-ph_med-mad-normalized-logratio.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/motrpac_pass1b-06_t53-cortex_prot-ph_med-mad-normalized-imputed-logratio.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/motrpac_pass1b-06_t53-cortex_prot-ph_normalized-data-feature-annot.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/motrpac_pass1b-06_t53-cortex_prot-pr_med-mad-normalized-logratio.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/motrpac_pass1b-06_t53-cortex_prot-pr_med-mad-normalized-imputed-logratio.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/motrpac_pass1b-06_t53-cortex_prot-pr_normalized-data-feature-annot.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/motrpac_pass1b-06_t55-gastrocnemius_prot-ph_med-mad-normalized-logratio.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/motrpac_pass1b-06_t55-gastrocnemius_prot-ph_med-mad-normalized-imputed-logratio.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/motrpac_pass1b-06_t55-gastrocnemius_prot-ph_normalized-data-feature-annot.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/motrpac_pass1b-06_t55-gastrocnemius_prot-pr_med-mad-normalized-logratio.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/motrpac_pass1b-06_t55-gastrocnemius_prot-pr_med-mad-normalized-imputed-logratio.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/motrpac_pass1b-06_t55-gastrocnemius_prot-pr_normalized-data-feature-annot.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/motrpac_pass1b-06_t58-heart_prot-ph_med-mad-normalized-logratio.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/motrpac_pass1b-06_t58-heart_prot-ph_med-mad-normalized-imputed-logratio.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/motrpac_pass1b-06_t58-heart_prot-ph_normalized-data-feature-annot.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/motrpac_pass1b-06_t58-heart_prot-pr_med-mad-normalized-logratio.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/motrpac_pass1b-06_t58-heart_prot-pr_med-mad-normalized-imputed-logratio.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/motrpac_pass1b-06_t58-heart_prot-pr_normalized-data-feature-annot.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/motrpac_pass1b-06_t58-heart_prot-ac_med-mad-normalized-logratio.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/motrpac_pass1b-06_t58-heart_prot-ac_med-mad-normalized-imputed-logratio.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/motrpac_pass1b-06_t58-heart_prot-ac_normalized-data-feature-annot.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/motrpac_pass1b-06_t58-heart_prot-ub_med-mad-normalized-logratio.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/motrpac_pass1b-06_t58-heart_prot-ub_med-mad-normalized-imputed-logratio.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/motrpac_pass1b-06_t58-heart_prot-ub_normalized-data-feature-annot.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/motrpac_pass1b-06_t59-kidney_prot-ph_med-mad-normalized-logratio.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/motrpac_pass1b-06_t59-kidney_prot-ph_med-mad-normalized-imputed-logratio.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/motrpac_pass1b-06_t59-kidney_prot-ph_normalized-data-feature-annot.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/motrpac_pass1b-06_t59-kidney_prot-pr_med-mad-normalized-logratio.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/motrpac_pass1b-06_t59-kidney_prot-pr_med-mad-normalized-imputed-logratio.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/motrpac_pass1b-06_t59-kidney_prot-pr_normalized-data-feature-annot.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/motrpac_pass1b-06_t66-lung_prot-ph_med-mad-normalized-logratio.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/motrpac_pass1b-06_t66-lung_prot-ph_med-mad-normalized-imputed-logratio.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/motrpac_pass1b-06_t66-lung_prot-ph_normalized-data-feature-annot.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/motrpac_pass1b-06_t66-lung_prot-pr_med-mad-normalized-logratio.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/motrpac_pass1b-06_t66-lung_prot-pr_med-mad-normalized-imputed-logratio.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/motrpac_pass1b-06_t66-lung_prot-pr_normalized-data-feature-annot.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/motrpac_pass1b-06_t68-liver_prot-ph_med-mad-normalized-logratio.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/motrpac_pass1b-06_t68-liver_prot-ph_med-mad-normalized-imputed-logratio.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/motrpac_pass1b-06_t68-liver_prot-ph_normalized-data-feature-annot.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/motrpac_pass1b-06_t68-liver_prot-pr_med-mad-normalized-logratio.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/motrpac_pass1b-06_t68-liver_prot-pr_med-mad-normalized-imputed-logratio.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/motrpac_pass1b-06_t68-liver_prot-pr_normalized-data-feature-annot.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/motrpac_pass1b-06_t68-liver_prot-ac_med-mad-normalized-logratio.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/motrpac_pass1b-06_t68-liver_prot-ac_med-mad-normalized-imputed-logratio.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/motrpac_pass1b-06_t68-liver_prot-ac_normalized-data-feature-annot.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/motrpac_pass1b-06_t68-liver_prot-ub_med-mad-normalized-logratio.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/motrpac_pass1b-06_t68-liver_prot-ub_med-mad-normalized-imputed-logratio.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/motrpac_pass1b-06_t68-liver_prot-ub_normalized-data-feature-annot.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/motrpac_pass1b-06_t70-white-adipose_prot-ph_med-mad-normalized-logratio.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/motrpac_pass1b-06_t70-white-adipose_prot-ph_med-mad-normalized-imputed-logratio.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/motrpac_pass1b-06_t70-white-adipose_prot-ph_normalized-data-feature-annot.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/motrpac_pass1b-06_t70-white-adipose_prot-pr_med-mad-normalized-logratio.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/motrpac_pass1b-06_t70-white-adipose_prot-pr_med-mad-normalized-imputed-logratio.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/motrpac_pass1b-06_t70-white-adipose_prot-pr_normalized-data-feature-annot.txtSave normalized, batch corrected, and protein-corrected data. Only applicable for PTM datasets.
#Perform DEA only on PTM data
ptm_datasets <- grep("prot-pr",names(proteomics_data),value = T,invert = T)
for(dataset_name in ptm_datasets){
#Normalized, batch corrected, and protein corrected
x_prot_cor = proteomics_data[[dataset_name]][["norm_filtered"]][["norm_data_b_protein_corrected"]]
#Get ome and tissue values from dataset name
arr = strsplit(dataset_name,split=",")[[1]]
curr_ome = tolower(arr[2])
curr_t = tolower(arr[1])
out_b = paste0(out_bucket,curr_ome,"/data/")
#Add bid and pid
x_bids = pheno_data$viallabel_data[colnames(x_prot_cor),"bid"]
x_pids = pheno_data$viallabel_data[colnames(x_prot_cor),"pid"]
newx_prot_cor = x_prot_cor
for(j in 1:ncol(newx_prot_cor)){newx_prot_cor[,j] = round(as.numeric(newx_prot_cor[,j]),digits=5)}
newx_prot_cor = rbind(x_bids,newx_prot_cor)
newx_prot_cor = rbind(x_pids,newx_prot_cor)
newx_prot_cor = rbind(colnames(x_prot_cor),newx_prot_cor)
colnames(newx_prot_cor) = NULL
rownames(newx_prot_cor)[1:3] = c("viallabel","pid","bid")
# save the output to the target bucket
local_fname = file.path(local_data_dir,
paste0("motrpac_pass1b-06_",
curr_t,"_",curr_ome,
"_med-mad-normalized-protein-corrected-logratio.txt"))
message("Write Local: ", local_fname)
write.table(newx_prot_cor,
file=local_fname,
sep="\t",
quote=FALSE,
row.names = TRUE,
col.names = FALSE)
if(upload_to_bucket){
cmd = paste(gsutil_cmd,"cp",local_fname,out_b)
message("Copy to Bucket: ", cmd)
system(cmd)
}
}## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/motrpac_pass1b-06_t53-cortex_prot-ph_med-mad-normalized-protein-corrected-logratio.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/motrpac_pass1b-06_t55-gastrocnemius_prot-ph_med-mad-normalized-protein-corrected-logratio.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/motrpac_pass1b-06_t58-heart_prot-ph_med-mad-normalized-protein-corrected-logratio.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/motrpac_pass1b-06_t58-heart_prot-ac_med-mad-normalized-protein-corrected-logratio.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/motrpac_pass1b-06_t58-heart_prot-ub_med-mad-normalized-protein-corrected-logratio.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/motrpac_pass1b-06_t59-kidney_prot-ph_med-mad-normalized-protein-corrected-logratio.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/motrpac_pass1b-06_t66-lung_prot-ph_med-mad-normalized-protein-corrected-logratio.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/motrpac_pass1b-06_t68-liver_prot-ph_med-mad-normalized-protein-corrected-logratio.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/motrpac_pass1b-06_t68-liver_prot-ac_med-mad-normalized-protein-corrected-logratio.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/motrpac_pass1b-06_t68-liver_prot-ub_med-mad-normalized-protein-corrected-logratio.txt## Write Local: ~/github/MoTrPAC/motrpac-bic-norm-qc/data/proteomics/DF_PASS_20210731/motrpac_pass1b-06_t70-white-adipose_prot-ph_med-mad-normalized-protein-corrected-logratio.txt